A new Australian study published Tuesday finds that AI Breast Cancer Detection could help identify women at higher risk of even after clear mammograms, potentially improving early detection and saving lives.
AI Tool Helps Identify Hidden Cancer Risk After Clear Mammograms
Researchers report that an AI system called BRAIx can analyze mammograms to predict which women may develop breast cancer within the next four years, even after screening results appear normal.
The study, published in the medical journal The Lancet Digital Health, examined screening data from more than 95,000 Australian women and more than 4,000 women in Sweden.
Among the Australian participants, about 1.1 percent, or 1,098 women. Developed breast cancer within four years after receiving a clear mammogram. In the Swedish group, 6.9 percent developed cancer within two years of a negative screening.
Researchers found the AI system effectively flagged women who later developed cancer, particularly within one to two years of a clear mammogram.
“AI tools such as BRAIx may help identify women who could benefit from additional testing earlier,” researchers wrote in the study.
Early detection is critical. In Australia, more than 20,000 women are diagnosed with breast cancer each year, and about 3,300 die from the disease.
Screening Program Faces Challenges With Interval Cancers
Australia’s national screening program, BreastScreen Australia, offers free mammograms every two years to women aged 50 to 74. The program began in 1992 and currently attracts participation from just over half of eligible women.
Despite screening, roughly 25 percent of cancers are diagnosed between routine screenings. These cases, known as interval cancers, are often aggressive and harder to treat.
Experts say one reason is breast density, a measure of glandular tissue in the breast. Higher breast density can increase cancer risk and make tumors harder to detect on mammograms.
Current risk assessments often rely on questionnaires, family history, and genetic testing. Genetic tests commonly look for mutations in genes such as BRCA1 and BRCA2, which significantly raise lifetime breast and ovarian cancer risk.
AI Breast Cancer Detection analyzes imaging patterns directly, offering another way to estimate risk.
“These technologies could complement existing screening methods,” researchers said, noting that women flagged as higher risk could receive follow-up scans such as MRI or contrast-enhanced mammography.
Researchers Call For Broader Trials And National Review
The study also highlighted limitations. BRAIx predicts cancer risk within a four-year window, making direct comparisons with lifetime genetic risk tests difficult.
Researchers also noted that the breast density data used in the analysis came from a different tool than the one currently used by the national screening program.
Still, the findings add to growing evidence supporting AI-assisted screening. A 2024 Swedish study using AI-based risk assessment referred about 7% of women for additional MRI scans. Among them, about 6.5 percent were found to have cancers missed during routine mammograms.
BRAIx is currently being tested in clinical settings through the BreastScreen Victoria program, where it assists radiologists in interpreting mammograms.
Health experts say the technology could eventually support a more personalized approach to screening.
Researchers suggest Australia may benefit from a national review of AI tools in breast screening programs.
If implemented carefully, they say AI Breast Cancer Detection could help detect cancers earlier and reduce breast cancer deaths.
