Deep learning model that determines the imaging biomarkers on screening mammograms, to predict a patient's risk for developing breast cancer, was developed by researchers at Massachusetts General Hospital (MGH), presented at the annual meeting of the Radiological Society of North America (RSNA).
This model has much greater accuracy than traditional risk assessment tools. "Every woman's mammogram is unique to her just like her thumbprint. It contains imaging biomarkers that are highly predictive of future cancer risk, but until we had the tools of deep learning, we were not able to extract this information to improve patient care", said senior author Constance D. Lehman, M.D., Ph.D., division chief of breast imaging at MGH.
Deep learning model in breast cancer:
On comparing the commercially available risk assessment model (Tyrer-Cuzick version 8) with the deep learning model for predicting the future breast cancer within five years, the latter achieved a predictive rate of 0.71, significantly outperforming the traditional risk model, which achieved a rate of 0.61. This deep learning algorithm was able to translate the full diversity of subtle imaging biomarkers in the mammogram.
"Traditional risk models can be time-consuming to acquire and rely on inconsistent or missing data. A deep learning image-only risk model can provide increased access to more accurate, less costly risk assessment and help deliver on the promise of precision medicine", said Leslie Lamb, M.D., M.Sc., breast radiologist at MGH.
Breast Cancer Facts and Statistics:
- About 12% will develop invasive breast cancer over the course of her lifetime.
- About 42,170 women in the U.S. are expected to die in 2020 from breast cancer.
- Key risk factors for breast cancer are sex (being a woman) and age (growing older).
- Breast cancer is the second leading cause of cancer death in women.