AI for Breast Cancer Risk Prediction Beyond Breast Density.
Authors
Abstract
AI has been increasingly adopted for breast cancer screening to improve cancer detection. Its role in mammography has expanded from cancer detection to image-based risk prediction. Recent advances have enabled AI to integrate subtle early imaging signs of breast cancer and mammographic patterns beyond breast density. Since the introduction of Mirai, an AI model designed to predict breast cancer risk within five years, research on mammography-based risk prediction has been growing. Multiple studies have demonstrated that AI models achieve superior performance compared to traditional clinical risk models across diverse screening populations. This advantage is reported to be greater in advanced breast cancers. An accurate risk assessment enables the identification of women who may benefit from supplemental breast cancer screening. This review provides an overview of the applications of AI in mammography in the era of breast cancer risk prediction, its medical evidence to date, and suggestions for future directions.