FDA authorizes 1st AI tool to predict 5-year breast cancer risk from routine mammograms

The U.S. Food and Drug Administration has OK’d the first artificial intelligence tool to predict five-year breast cancer risk from routine mammograms, the product’s developer announced late Friday. 

The de novo authorization is for Clairity Breast, from the Boston-based vendor of the same name. Most risk assessment models rely on age and genetics, the company noted, but 85% of women diagnosed with the disease do not have a family history. Plus, half of those with breast cancer have no identifiable risk factors. 

Clairity analyzes screening mammograms and can detect “subtle imaging features” correlated with long-term risk. This makes early risk prediction feasible “based on a screening mammogram alone.” The software can produce a validated five-year risk score, which is delivered to radiologists through “existing clinical infrastructures.” 

"Clairity’s FDA authorization is a turning point for more women to access the scientific advances of AI-driven cancer risk prediction,” Larry Norton, MD, founding scientific director of the Breast Cancer Research Foundation and a breast oncologist with Memorial Sloan Kettering, said in a statement May 30. "Breast cancer is rising, especially among younger women, yet most risk models often miss those who will develop the disease. Now we can ensure more women get the right care at the right time."

Clairity was founded in 2020 by Constance “Connie” Lehman, MD, PhD, a Harvard professor of radiology and former chief of breast imaging at Massachusetts General (where she still works). It also is backed by life sciences investment firm Santé Ventures and ACE Global Venture, a business advisory and investment banking company. Clairity cited previous research estimating the global breast cancer prediction market is worth $63 billion.

Lehman and colleagues said the system analyzes breast images “at the pixel level” to identify those who will likely develop cancer. 

"Advancements in AI and computer vision can uncover hidden clues in the mammograms—often invisible to the human eye—to help predict future risk," Lehman said. 

Clairity Breast has been trained across “millions of mammogram images,” matched with five-year outcome data to help radiologists identify patterns in breast tissue that correlate with future cancer development. The company hopes for commercial launch of Clairity Breast by the end of 2025. 

“What makes the availability of Clairity Breast a true sea change is that we’re now anticipating cancer from patterns in breast tissue, in an otherwise normal screening, before it’s even there,” CEO Jeff Luber, MBA, said in the announcement. “Clairity Breast is designed to fit seamlessly into current clinical infrastructure to help providers scale precision prevention—with the goal of reducing late-stage diagnoses, lowering costs and saving more lives.”

While the FDA has cleared AI-powered tools for breast cancer detection, contributing to risk management decisions, the agency has not yet OK’d one for predicting five-year risk from mammograms. Imaging center operator RadNet Inc. recently agreed to acquire breast imaging AI vendor iCAD earlier this year. During a call with investors, leaders discussed the development of new AI product that allows radiologists to predict a woman’s potential of developing the disease based on images. A recent study in RSNA’s Radiology also explored five-year cancer risk prediction using negative mammograms. You can find more about Clairity’s previously published research here.