Constance D. Lehmen, MD, PhD, Professor of Radiology at Harvard Medical School and the senior author of the study, stated that conventional methods for assessing breast cancer risk—such as age, family history, genetics, and breast tissue density—were outperformed by the AI model. This research was presented at the annual meeting of the Radiological Society of North America (RSNA).
The Clairity Breast system is the first such model to receive clearance from the U.S. Food and Drug Administration (FDA). It was trained using a massive dataset comprising 421,499 mammograms originating from 27 clinical sites across Europe, South America, and the United States.
The researchers incorporated images from both healthy individuals and those diagnosed with the disease, enabling the model to learn subtle features within breast tissue structure that correlate with future disease risk. The development team points out that Clairity Breast is capable of detecting tissue alterations that are imperceptible during routine physical examinations or standard image review.
The model was validated using 236,422 digital mammograms sourced from five U.S. institutions, alongside 8,810 mammograms from a European facility. The data spanned collection periods from 2011 through 2017. Additionally, information regarding breast density and five-year cancer incidence records from patient charts were integrated into the evaluation.
The risk estimates generated by the AI were categorized based on guidelines established by the National Comprehensive Cancer Network (NCCN): low (<1.7%), intermediate (1.7–3.0%), and high (>3.0%).
When factoring in breast tissue density, the AI-calculated high-risk group developed breast cancer at a rate more than four times higher than the low-risk group (5.9% incidence versus 1.3%). In contrast, risk stratification based solely on traditional breast density assessment yielded only a marginal difference (3.2% for dense tissue compared to 2.7% for less dense tissue).
According to the World Health Organization (WHO), in 2022, breast cancer was the most frequently diagnosed cancer among women in 157 out of 185 countries. That same year, the disease was newly diagnosed in 2.3 million women globally, leading to 670,000 fatalities worldwide.
Various figures suggest that around 54,000 women in Russia receive this diagnosis annually. Concurrently, only about 30% of these women attend annual mammologist appointments, 40% are examined less frequently, and 30% avoid seeing a doctor altogether.
About the Author
