Support the Adoption of a Revolutionary Deep Learning Mammography Model for Breast Cancer Risk Prediction.

1 in 8 women will develop breast cancer in her lifetime. However, predicting who will be affected remains a challenge. Until recently, methods for assessing risk, such as breast density, have not reliably stratified patient risk of developing breast cancer. As a result, current screening guidelines adopt a “one size fits many” approach.

In the last 5 years, a new generation of AI-based risk models has been developed. These methods analyze existing mammograms and accurately predict a patient’s risk of developing breast cancer in the next 5 years. MIRAI, an AI model developed by MIT and Mass General Brigham researchers, has been tested in large, diverse populations both in the United States and around the world. It has consistently demonstrated superior performance compared to all existing risk stratification methods. In contrast to traditional models, MIRAI’s performance is equitable across race, age, and breast densities.

Unfortunately, today, AI-based screening models are not yet used in clinical practice. This delay is partially due to the slow change in clinical guidelines and insurance policies. To address this, we have launched the #CancerInSight initiative aimed at accelerating the adoption of AI-driven risk stratification to benefit patients. Achieving this goal will require updates to clinical guidelines and support from insurance companies.