The AI detects tumors more often and earlier in the Dutch breast cancer screening program. These tumors can then be treated at an anterior stage. This has been demonstrated by researchers led by Radboud University Medical Center in a study published in Lancet Digital Health. The use of AI could reduce the workload and save millions of euros per year.
Previous research in Sweden has already shown that AI detects breast cancer on mammography more frequently than radiologists. In addition, AI can reduce the workload for radiologists. Now it seems that AI can also replace the second radiologist in the Dutch breast cancer screening program. This even leads to the detection of more tumors and an anterior stage, which is later revealed clinically significant.
AI is more likely to be right
The researchers, led by the Radboudumc Ritta Mann mammary radiologist, analyzed 42,000 breast scans. These mammograms were taken as part of the Dutch screening program in the Utrecht region. Traditionally, two radiologists examine these analyzes, as well as the standard practice of breast cancer screening. In this study, researchers also evaluated the analyzes using AI developed by MEDICAL Screenpoint. In addition, they followed women whose scans were analyzed for almost four and a half years, with several analyzes available for many of them.
The study has shown that a radiologist working with AI detects more tumors than two radiologists alone. Tumors are also identified earlier when AI is involved. “Sometimes AI detects a tumor that radiologists do not yet recognize as such. We call this a false positive. But often, this tumor appears in a later scan after all. Therefore, the AI was just earlier, ”explains Suzanne Van Winkel, a doctoral candidate. “As the radiologist stimulates the alarm, it often concerns larger invasive tumors, which certainly need treatment, as soon as possible. »»
Replace the second radiologist
In Sweden, AI is already used to analyze screening mammograms. “They replace the second radiologist with AI. It is only if the AI is uncertain, a second radiologist intervenes, ”explains Mann. “We see that radiologists work well with AI, which leads to more tumors detected without a significant increase in unnecessary monitoring checks for women. »»
This study shows that AI is also good enough in the Dutch framework to replace the second radiologist in the analysis of the screening scan. This could save several million euros per year. However, AI is not currently used in the Netherlands. “In Sweden, screening is organized regionally, while in the Netherlands, it is organized nationally. This makes the implementation of the more difficult Logistics AI here, ”explains Mann. “Computer infrastructure in the Netherlands is not yet ready. This requires funding.
