SBI Wrap Up and Newly Published Research
SBI/ACR Breast Imaging Symposium Wrap Up:
The 2022 SBI/ACR Breast Imaging Symposium was packed with excellent presentations on a range of topics, including particular focus on artificial intelligence (AI) in breast imaging, which improves both early detection and risk assessment, and the importance of addressing disparities in breast imaging. Black women, American Indians, and Alaska natives are significantly more likely to be diagnosed with breast cancers at younger ages and later stages than non-Hispanic white women and are more likely to die from breast cancer. The importance of improving access to breast cancer screening and earlier risk assessment can’t be stressed enough.
Congratulations to DB-i’s Chief Scientific Advisor Dr. Wendie Berg and Medical Advisory Board members Drs. Stamatia Destounis, Carrie Hruska, Emily Conant, Jean Seely, Regina Hooley, and Laurie Margolies on their outstanding presentations and continuing to move the field forward.
Newly Published:
Just out in the Science Translational Medicine, research article, A risk model for digital breast tomosynthesis to predict breast cancer and guide clinical care. Modified from the abstract: Screening with digital breast tomosynthesis (DBT) improves breast cancer detection and reduces false positives. However, currently, no breast cancer risk model takes advantage of the additional information generated by DBT imaging for breast cancer risk prediction. A DBT-based short-term risk model was developed and internally validated for predicting future late-stage and interval breast cancers after negative screening exams. The model’s ability to distinguish women who were diagnosed with interval or late-stage cancers one year after a negative mammogram from those who were not (discrimination performance of 1-year risk), was considered excellent (AUC = 0.82), and higher than models using FFDM or traditional risk factors. In simulation modeling, if 12% of the women at highest risk had been offered supplemental screening, then potentially up to 59% of the cancers may have been detected compared with 39% of cancers using FFDM and 24% using lifestyle-family–based risk models.