Using AI to read mammograms cut risk of developing aggressive breast cancer, study finds
A large-scale study conducted in Sweden suggests that incorporating artificial intelligence (AI) into mammogram screenings could significantly reduce the rate of aggressive breast cancers. The research, involving over 100,000 women, found a 12% decrease in interval cancers – those detected between routine screenings – when AI assisted radiologists in interpreting the images.
The Promise of AI in Early Detection
Mammograms are a crucial tool in the early detection of breast cancer, allowing for more effective treatment. This study, published in The Lancet, demonstrates that AI can enhance the accuracy of these screenings. Researchers found that AI assistance led to fewer instances of aggressive and advanced breast cancers being diagnosed at a later stage.
The study compared two approaches: AI-supported mammography for half the participants and traditional “double reading” by two radiologists for the other half. Double reading is common in much of Europe, but is not the standard practice in Canada, where typically one radiologist reviews each mammogram.
Understanding Interval Cancers
Interval cancers are particularly concerning because they are often more aggressive and have a poorer prognosis. The Swedish study specifically tracked the incidence of these cancers, finding a notable reduction with the implementation of AI screening. Dr. Kristina Lång, the lead author, described the 12% decrease as “a big sigh of relief.”
Potential Impact on Healthcare Systems
In 2024, approximately 30,500 Canadians were diagnosed with breast cancer, with an 89% five-year survival rate. AI-assisted mammography could potentially improve these outcomes. Dr. Jean Seely, head of breast imaging at the Ottawa Hospital, noted that treatment for stage 3 interval cancers can cost between $100,000 and $200,000 per patient. Therefore, reducing the number of these advanced diagnoses could also lead to significant cost savings for healthcare systems.
The study utilized a triage system, where AI identified low-risk mammograms for review by a single radiologist and flagged high-risk cases for review by two. This approach could optimize resources and focus expertise where it’s most needed.
What Might Come Next
If these findings are replicated in other populations and healthcare settings, we could see a wider adoption of AI-assisted mammography. Hospitals and clinics may begin investing in AI software to support their radiology departments. Further research will be needed to determine the long-term impact on mortality rates and to assess the cost-effectiveness of this technology. It’s also possible that guidelines for breast cancer screening could be updated to reflect the benefits of AI assistance.
Frequently Asked Questions
What is an interval cancer?
An interval cancer is an invasive tumor that appears between routine mammogram screenings. They can be harder to detect and are often more aggressive.
How does AI assist in mammogram readings?
The AI software analyzes mammograms and flags areas that may indicate abnormalities, assisting radiologists in their interpretation.
Is AI-assisted mammography currently used in Canada?
While not yet standard practice, some Canadian hospitals, like the Ottawa Hospital, are using research funds to purchase AI software to help read mammograms.
As AI technology continues to evolve, how might it reshape the future of preventative healthcare and early disease detection?