Computational Breast Imaging Group - Research

Computer-Assisted Risk Estimation (CARe) for Breast Cancer

Identifying women at high risk of breast cancer is critical for preventing breast cancer. Women at high risk can benefit from screening with breast MRI which has been shown to detect breast cancers that are not detected by mammography. High-risk women can also benefit from risk-reduction therapy with chemoprevention drugs which can reduce the incidence of breast cancer by 48%. Unfortunately, no method currently exists to accurately identify high-risk women from the general population. Most of the research has focused on women at increased familial risk (i.e. women with certain gene mutations), which only account for the 5%-10% of the breast cancers in the general population. On the other hand, the National Cancer Institute's (NCI) breast cancer risk assessment tool for the general population is not accurate in predicting which individual women are most likely to develop breast cancer. For example, it is estimated that of the 10 million US women eligible by NCI's risk assessment guidelines for chemoprevention drugs, only about 25% would actually benefit, exposing a considerable fraction of the population to side effects such as stroke, endometrial cancer, and cataracts.

One method to improve breast cancer risk estimation is to also consider breast density. Breast density is the strongest known risk factor for breast cancer after age. Currently, breast density is not routinely included in the breast cancer risk assessment calculations in clinical practice. Studies have shown the potential to improve NCI's risk assessment tool by including measures of breast density. However, these improvements have been minimal and not widely implemented due to the lack of automated methods to accurately measure breast density. Digital mammography is a new breast imaging modality increasingly replacing conventional screen-film mammography in breast cancer screening. Digital mammography offers the opportunity to measure breast density more accurately using automated methods. Our project aims to improve breast cancer risk estimation by developing a new Computer-Assisted Risk Estimation (CARe) tool for breast cancer that will incorporate measures of breast density from digital mammography. Our hypothesis is that a risk assessment model that includes measures of breast density with a woman's family history and additional clinical breast cancer risk factors can be more accurate than NCI's current model and improve breast cancer risk estimation for women.

Our study holds the promise to improve the current standards in breast cancer prevention. We envision providing a unique setting in which breast cancer risk assessment and patient education can be combined to empower women with knowledge about their personal risk and providing a much needed fully-automated risk assessment tool for physicians. The information can be used to identify women at high risk, so that tailored screening recommendations, surveillance, and risk reduction therapies can be offered more effectively.