A study conducted by RadNet subsidiary DeepHealth revealed that more than a third of women across 10 health care practices opted into a self-pay, artificial intelligence (AI)-enhanced breast cancer screening program — and enrollees were 21% more likely to have cancer detected.
The study, presented at the RSNA meeting, evaluated the use of AI in screening mammography as part of a self-pay option. AI technology served as an additional layer of review, alongside radiologists, to support decision-making and risk prediction. Enrolled patients also benefited from a "safeguard review," where expert breast radiologists provided an additional assessment in cases of disagreement between the AI and initial human review.
Out of 747,604 women who underwent screening mammography during the study's 12-month period, cancer detection rates were 43% higher on average for participants in the AI-enhanced program compared to unenrolled patients. Of this increase, 21% was attributed directly to the AI component, while the remaining 22% was linked to higher enrollment rates among high-risk patients.
“This is the first report on results from a program that provides an AI-powered enhanced review that patients can elect to enroll in,” said Bryan Haslam, the study’s lead author. “The number of women electing for this program is now at 36% and growing, and the rate of cancer detection continues to be substantially higher for those women.”
The study also noted a 21% higher recall rate among enrolled patients, meaning they were more frequently called back for further imaging. However, these recalls resulted in a 15% higher positive predictive value for cancer, demonstrating that more diagnoses were confirmed.
Researchers aim to further investigate the effectiveness of the AI-driven safeguard review through prospective randomized, controlled trials to reduce potential self-selection bias.
