NYU researchers develop AI tool for personalized breast cancer risk prediction

Researchers at NYU Langone Health and the Perlmutter Cancer Center have developed a deep-learning AI tool, named NYU-DRP, designed to predict a woman’s five-year risk of developing breast cancer. Unlike traditional risk models that rely on broad clinical factors such as age, family history, and reproductive history, this tool utilizes longitudinal digital breast tomosynthesis (3D mammograms) taken over multiple years.

In a study published in the American Journal of Roentgenology, NYU-DRP demonstrated superior predictive accuracy compared to other models. The tool correctly identified high-risk individuals 72 percent of the time, outperforming single 3D mammogram analysis (70 percent) and AI-assisted 2D testing (68 percent).

By analyzing how breast tissue changes across multiple screenings over time, the model offers a more personalized approach to breast cancer screening. This advancement aims to move beyond current clinical calculators, such as the Gail and Tyrer-Cuzick models, by leveraging detailed imaging information to better stratify patient risk.