A Deep Learning Breast Cancer Risk Model for Precise Supplemental Screening

1

The FDA requires imaging facilities to notify patients about breast density, which affects cancer risk and mammography accuracy.
2

Current policies for supplemental screening based on breast density may lead to overuse of resources and increased false positives.
3

Advanced deep learning models can estimate breast cancer risk from mammograms, outperforming traditional risk assessment methods.
4

The study analyzed 123,091 mammograms to compare a deep learning model's risk predictions with traditional breast density assessments.
5

The deep learning model, Mirai, generates a numerical risk score without relying on subjective assessments or patient questionnaires.

1