AI-supported mammography screening results in fewer aggressive and advanced breast cancers
A groundbreaking, randomized controlled trial involving over 100,000 women in Sweden has demonstrated that artificial intelligence (AI)-supported mammography not only improves cancer detection during screening but also reduces breast cancer diagnoses in the years following a screening appointment by 12%. The findings, published in The Lancet, represent a significant step forward in the fight against breast cancer.
The Promise of AI in Early Detection
The study, known as the MASAI trial, builds on earlier research indicating the potential of AI in mammography. Interim results, published in The Lancet Oncology in 2023, showed a 44% reduction in the workload for radiologists reviewing screenings. A separate early analysis in The Lancet Digital Health revealed a 29% increase in cancer detection without a corresponding increase in false positives.
Reducing Interval Cancers
Despite current screening practices, a concerning percentage of breast cancers – estimated between 20% and 30% – are diagnosed *after* a negative mammogram and before the next scheduled screening. These “interval cancers” are often more aggressive and harder to treat. The Swedish trial specifically investigated whether AI-supported mammography could reduce the occurrence of these challenging cases.
How the Trial Worked
From April 2021 to December 2022, women participating in routine mammography screenings at four Swedish sites were randomly assigned to either AI-supported screening or standard double reading by radiologists. In the AI-supported arm, a specialized AI system analyzed the images, flagging high-risk cases for double reading by radiologists and triaging low-risk cases for single reading. Radiologists also utilized the AI as a detection aid, with the system highlighting potentially suspicious areas.
Significant Results Across Key Metrics
The results showed a 9% increase in the detection of cancer cases at screening in the AI-supported group (81% vs. 74%). Importantly, the rate of false positives remained comparable between the two groups (1.5% vs. 1.4%). Furthermore, the AI group experienced a 16% reduction in invasive cancers, a 21% reduction in large cancers, and a 27% reduction in aggressive subtype cancers compared to the control group.
Looking Ahead
Dr. Kristina Lång of Lund University, Sweden, cautioned that introducing AI into healthcare requires careful consideration and continuous monitoring to ensure consistent performance across different screening programs. Further research is planned to assess the long-term benefits, cost-effectiveness, and potential variations in outcomes over time.
The AI system used in the trial was rigorously trained and validated using over 200,000 examinations from institutions across more than ten countries.
Frequently Asked Questions
What was the primary goal of the MASAI trial?
The primary goal of the MASAI trial was to investigate whether the use of AI in breast cancer screening could improve the early detection of clinically relevant breast cancers and reduce the number of interval cancers.
Did the AI system replace radiologists in the trial?
No, the AI system did not replace radiologists. Radiologists still performed the screen readings, but they were supported by the AI, which highlighted suspicious findings and triaged cases for single or double reading.
What limitations did the authors identify in their study?
The authors noted that the study was conducted in a single country (Sweden), used only one type of mammography device and one AI system, and did not collect data on race and ethnicity, which may limit the generalizability of the results.
As AI technology continues to evolve, will this approach to breast cancer screening become more widespread, and what impact might that have on patient outcomes?