1 minute read

Mammography AI Can Help Predict Breast Cancer Risk

Artificial intelligence trained to read mammograms was better at predicting a woman’s future breast cancer risk than a commonly used risk assessment tool, a new Kaiser Permanente study found.

“We need to have better ways to assess a woman’s future breast cancer risk and artificial intelligence may be able to help, ” said study lead author Vignesh Arasu, MD, PhD, a research scientist at the Kaiser Permanente Division of Research and radiologist specializing in breast imaging at Kaiser Permanente Vallejo Medical Center. “Our study found that the artificial intelligence can see changes in breast tissue on a woman’s mammogram that can help us assess her risk of being diagnosed with cancer in the next 5 years. ”

The study published in Radiology included 324,000 women who had a mammogram at Northern California Kaiser Permanente facilities in 2016 that found no sign of breast cancer. Over the next 5 years, 4,584 of the women had a breast cancer diagnosis.

The researchers had the artificial intelligence read the women’s mammograms from 2016 and produce a 5-year breast cancer risk score. They also had the artificial intelligence read the mammograms and produce 5-year risk scores for 13,435 women who did not have a breast cancer diagnosis.

Then, the researchers used a breast cancer risk calculator to create a 5-year breast cancer risk score for both groups of women using information from their medical records at the time of their 2016 mammogram. The calculator creates a risk scored based on a woman’s age, race or ethnicity, first-degree family history of breast cancer, number of prior benign breast biopsies, and mammographic breast density.

The researchers found that the artificial intelligence was better than the risk calculator at predicating which women would be diagnosed with breast cancer over the next 5 years. The team said this suggests that artificial intelligence used alone or with current risk prediction tools could help doctors develop personalized breast cancer screening recommendations for women.

“Right now, we mostly use artificial intelligence to help radiologists identify a visible cancer, ” said Dr. Arasu. “But these algorithms can also see changes in breast tissue in addition to breast density that tell us there are changes happening in the breast tissue that can be used to predict risk. ”

For more information on how Kaiser Permanente is transforming health through research, visit divisionofresearch.kaiserpermanente.org.

This article is from: