A breast screening program in the United States is showing promising results when using artificial intelligence (AI) in mammograms, which are important tests that help detect breast cancer. Women who paid for these AI-enhanced scans were found to be 21% more likely to have cancer identified than those who did not use this technology.
The research was conducted by DeepHealth, a company focused on AI and connected to RadNet, a major radiology company. They shared their results during an important annual meeting for radiology professionals. The study involved 747,604 women who had mammograms performed within a year. It found that those who participated in this AI program had a 43% higher rate of cancer detection compared to those who didn’t take part.
Read More: China Launches Antitrust Investigation Against Nvidia
Notably, some women chose to pay extra for a special AI software that had been approved by the U.S. Food and Drug Administration. This software works like a second opinion for doctors, helping them notice anything unusual in the mammogram images that could indicate cancer.
However, it’s important to understand that part of the increase in the cancer detection rate—22%—may have been because women who are at higher risk for breast cancer were more likely to join this program in the first place. This is known as selection bias. Nevertheless, the research indicated that the remaining 21% of the increase in detection rates could be attributed to the AI technology assisting radiologists in identifying potential issues more effectively.
While these findings are encouraging, they are not definitive. The researchers are now looking to conduct more rigorous studies, known as randomized controlled trials, to more accurately determine the real benefits of using AI in mammogram screenings.
