Many women interested in AI-based risk assessment for breast cancer

Aug 10, 2026

A new study found that 80.5% of women are comfortable with AI-based breast cancer risk assessment using mammograms, and most prefer receiving their risk results from their clinical providers. Women found to be at increased risk showed high interest in additional screening and genetic testing, with over half willing to pay out of pocket for these services.

80.5% of surveyed women reported comfort with AI-based breast cancer risk assessment, compared to only 31% who preferred using personal or family history
Women strongly prefer receiving risk results from their own clinical providers—gynecologists, radiologists, or primary care physicians—rather than automated systems
90.3% of women found to be at increased risk expressed interest in additional screening tests
72% of high-risk women showed interest in genetic testing, versus 50% of women at average risk

Many women are comfortable enough with AI-based breast cancer risk assessment using their screening mammogram, according to survey findings published August 8 in Clinical Imaging.

Researchers led by Liliana Light, an MD student from Howard University in Washington, D.C. also found that women prefer receiving their breast cancer risk information from their clinical providers and that finding higher breast cancer risk is tied to high interest by women in additional screening.

“The growing acceptance of AI-based models, genetic testing, and supplemental imaging tools offers the potential to enhance breast cancer detection and ultimately improve patient outcomes,” Light and co-authors wrote.

Accurate assessment of women’s risk for developing breast cancer is needed to determine which women should undergo supplemental imaging with whole breast ultrasound or MRI. AI-based risk models are being studied to assess breast cancer risk from mammograms. The researchers noted a lack of data on patient comfort with implementation of an automated risk assessment tool using mammography data.

The Light team evaluated women’s views on breast cancer, risk assessment, and screening tools. It also assessed women’s willingness to undergo or pay for supplementary screening or genetic testing based on their assessed risk.

The study included 319 surveys completed by patients at four medical centers. Clinics included radiology (n = 160), gynecology (n = 80), and internal medicine (n = 79).

The researchers reported the following findings:

Most respondents (80.5%) said they would be comfortable with AI-based breast cancer risk assessment, compared to 31.0% who preferred using personal/family history from a survey.

Women surveyed at gynecology and radiology offices preferred obtaining their risk results from their gynecologist (82.5%, p ≤ 0.001) or radiologist (57.3%, p = 0.006), while most women surveyed at the internal medicine clinic (63.3%, p = 0.081) preferred their primary care physician.

For women found to be at increased risk, 90.3% said they were interested in additional screening tests.

Most women at increased breast cancer risk (72%) showed interest in genetic testing, compared with 50% of women not at increased risk (p = 0.001).

Also, over half of women expressed a willingness to pay out of pocket for genetic testing, even when insurance coverage is unavailable. About 54% said they would be willing to pay up to $250.

“This finding is particularly relevant as genetic testing becomes more affordable and widely available in clinical settings,” the researchers wrote.

The study authors noted that women’s preferences for receiving their risk results “underscores the importance” of provider-patient relationships in risk communication.

“Future research should include more diverse populations and further examine factors influencing willingness to pay for supplemental screening, including perceived value, insurance coverage, and understanding of test benefits and risks, to support patient-centered integration of emerging screening strategies into clinical practice,” the authors wrote.

Read the full study here.