By: Jordan Poltorak
DAVIS, CA–The University has received a 15 million dollar grant from the National Cancer Institute to develop artificial intelligence (AI) that can detect breast cancer and create a risk model.
Mammograms are the most common way to diagnose breast cancer, but they are not without flaws. They can miss one in eight cases and give false-negative results.
This grant will go towards three different projects. Each of these projects works towards streamlining, perfecting the diagnosing process, and creating equity in cancer screening.
The first project will use AI to predict if women are at significant risk for developing advanced cancer. The risk models “include imaging features and evaluate FDA-approved AI scores from five vendors.”
Then, researchers will compare the AI results to those from a mammogram. Thus, improving the AI’s effectiveness and allowing women without a history of breast cancer to see if they are at risk.
The next project will work to reduce the inequities in the breast cancer screening process. The goal is to use AI algorithms to improve cancer detection and see if these are comparable to visiting a specialist.
The latter aspect is crucial for women from lower-income communities that may not have access to breast imaging specialists. Factors such as location, race, and income are all causes of variance in rates of breast cancer diagnosis.
Affluent white women have “historically benefited from screening mammography to detect breast cancer at earlier stages” which gives them a higher survival rate.
In the final project, researchers will attempt to close the gap between second breast cancer diagnoses. After initial treatment, women return for a yearly mammogram, where the second diagnosis often occurs.
However, this project would work to identify high-risk individuals and catch returning cancer sooner rather than later. The project covers high-risk factors such as age and breast density, with calculations of the chance signs that may be missed by a mammogram.
Those odds speak to why AI development is necessary. Researchers hope it will be more effective than a mammogram and accessible to all women.
A national research team led by Diana Miglioretti, a UC Davis Professor and Division Chief of Biostatistics, is conducting the experiments. Miglioretti has been a professor at UC Davis since 2013 and has been the Division Chief of her department since 2019.
Most of Miglioretti’s collaborative research is in breast cancer screening and radiation exposure as a result of medical imaging.
The National Cancer Institute is the “federal government’s principal agency for cancer research and training.” They receive funds from Congress and use them to fund cancer research at universities, hospitals, and laboratories.
Breast cancer is the second leading cause of death for women in the US. This grant will further research in the area and help women to live longer and access more equitable care.
