A recent study in Germany has found that Artificial Intelligence (AI) can improve breast cancer detection and reduce the workload on physicians. The research, published in the journal Nature Medicine, involved over 461,000 women who were part of a national screening program for breast cancer in Germany.
Unlike previous studies that used AI on existing data after patients had already been diagnosed, this study used AI prospectively on women between the ages of 50-69 who had no symptoms of breast cancer. The women were split into two groups, with one group having their screening data analyzed by two expert radiologists and the other group having at least one radiologist using an AI tool to assist in the analysis.
The results showed that the AI-assisted group had a 6.7% higher detection rate of breast cancer compared to the group analyzed solely by radiologists. When adjusting for potential variables, such as the age of the women diagnosed, this difference increased to a 17.6% higher detection rate in the AI-assisted group.
Study leader Alexander Katalinic, MD, stated that the study aimed to show that AI reporting is equivalent to human reporting, but the results showed that AI significantly improved the breast cancer detection rate. The study also found that both the AI-assisted and unassisted radiologists had a similar false-positive rate, which is important in reducing unnecessary distress for patients and costs for healthcare systems.
Stefan Bunk, CTO of the AI company Vara who designed the AI tool, believes that the study demonstrates the potential of AI to improve screening programs worldwide. The hope is that AI tools can eventually reduce the workload on doctors involved in cancer diagnosis, improve accuracy, and reduce false positives, ultimately improving the prognosis for women with breast cancer.
Overall, the study highlights the promising role that AI can play in improving breast cancer diagnosis and screening programs, potentially leading to better outcomes for patients in the future.