AI Model Leverages Mammogram Data to Improve Breast Cancer Risk Prediction

Currently, there is no method to determine who will develop breast cancer based on imaging alone, instead relying on a standard, questionnaire-based method that only assesses clinical risk factors such as age, race and family history. Researchers from Mallinckrodt Institute of Radiology (MIR) and WashU Medicine created a new AI model that significantly improves the accuracy of predicting the risk of breast cancer development using mammograms.

The algorithm detects subtle changes over time in repeated mammogram images that are not visible to the human eye, including breast density, texture and symmetry. It identified individuals at high risk of developing breast cancer 2.3 times more accurately than the standard method based on a total of 28,000 mammograms from two academic medical centers.

Debbie L. Bennett, MD, associate professor of radiology and chief of breast imaging, said their results indicate that the model is an exciting development in utilizing data from previous mammograms to help diagnose cancer earlier and guide clinical recommendations. “The prediction is never going to be perfect, but this study suggests the new algorithm is much better than our current methods.”

Read more about the study from WashU Medicine.