AI predicts breast cancer risk years before diagnosis, Norwegian study finds

Norwegian Institute of Public Health says AI use could improve early detection, reduce costs, and better target at-risk populations.

 A mobile breast screening truck. Seine-Maritime, France, 2023. (photo credit: Leitenberger Photography. Via Shutterstock)
A mobile breast screening truck. Seine-Maritime, France, 2023.
(photo credit: Leitenberger Photography. Via Shutterstock)

Scientists from the Norwegian Institute of Public Health (FHI), along with specialists from the University of California and the University of Washington, conducted a study published in October in the Journal of the American Medical Association Network Open. The research revealed that artificial intelligence (AI) can predict women's breast cancer risk years in advance by analyzing mammograms.

The study analyzed data from 116,495 women who participated in the Norwegian detection program between 2004 and 2018. Among these women, 1,607 were diagnosed with breast cancer. Using a commercially available AI algorithm, the researchers retrospectively assessed the mammograms, assigning a risk score from 0 to 100 indicating the likelihood of developing breast cancer.

"The algorithm was able to accurately predict which breast was at risk," the FHI announced, according to Euro2day. Notably, the breast that developed cancer received an AI score about twice as high as the other breast four to six years before diagnosis.

"We found that the breast that developed cancer received an AI score about twice as high as the other breast," explained Solveig Hofvind, project leader and director of the screening program at FHI, in a statement reported by BFMTV. "The study shows that AI algorithms already on the market can be used to develop more personalized screening programs."

By analyzing the AI scores, the researchers observed that the breast which eventually developed cancer consistently received higher risk scores compared to the other breast, as early as four to six years before diagnosis. This difference in scores underscores the potential of AI to detect subtle changes in mammograms that may be imperceptible to human radiologists.

In addition to enhancing early detection, the FHI highlighted other potential benefits of incorporating AI into breast cancer screening programs. The technology could lead to reduced costs by streamlining the screening process and enabling more efficient use of resources. Furthermore, by accurately identifying high-risk individuals, screening efforts can be better directed toward populations that would benefit the most, thereby maximizing the impact of preventive measures.

According to the World Health Organization (WHO), 670,000 women died of breast cancer in 2022, making it the most common form of cancer among women in most countries. Early detection is crucial in reducing mortality rates, and the study suggests that AI could be instrumental in achieving that goal.

Building on these findings, the Norwegian screening program launched a study involving 140,000 women last year to determine if AI can be as effective, or even better, than radiologists in diagnosing cancer. "The purpose of the project is to determine whether AI could be as efficient as, or even better than, radiologists in diagnosing cancer cases," Hofvind stated, according to Punch.

As AI continues to advance and integrate into various aspects of healthcare, its role in personalized medicine becomes increasingly evident. The ability to tailor screening programs based on individual risk profiles represents a shift toward more efficient and effective healthcare delivery.

Parallel to the Norwegian study, about 80,000 women in Sweden underwent mammogram screening for breast tumors between April 2021 and July 2022, as reported by ProTV News. These concurrent studies reflect a growing international interest in harnessing AI technology to combat breast cancer.


Stay updated with the latest news!

Subscribe to The Jerusalem Post Newsletter


The article was written with the assistance of a news analysis system.