AI Clarity Score Assesses Five-Year Breast Cancer Risk from Mammogram Views

An AI model analyzes four standard mammogram views to estimate a woman's likelihood of developing breast cancer over the next five years, Connie Lehman of Clarity, Inc. stated in an interview. The model takes inputs of two views from the right breast and two from the left breast, extracted from routine screening mammograms, and outputs a percentage score, Lehman explained.

She described this as a more targeted approach compared to traditional mammograms, which she said have limitations and are insufficient for personalized care. Lehman expressed interest in women's health and global health access, noting that she became excited about radiology due to the power of imaging the body.

"The power of the image, I thought it was so incredible and so untapped in healthcare, just the power of imaging the body, and the technology by which we can take images outside and inside the body," Lehman said.

The interview took place at the Imagination in Action event on April 9 and 10, where Boston Globe reporter Aaron Pressman spoke with Lehman. Artificial intelligence is used in radiology to help clinicians read scans, providing a foundation for such advancements. Lehman referenced early research on this type of technology in the 1970s.

She highlighted cases like an MIT professor who was diagnosed with breast cancer and was shocked by how little science was used to determine her next steps after diagnosis. A broadcast professional also battled breast cancer, despite an absence of family medical history, Lehman noted.

"She was shocked by how little science was being used to determine exactly what she should do next," Lehman said of the professor's experience.

Lehman pointed to broader challenges in health research, including NIH studies that exclude female mice because their cycles mess up the data. "The mammogram isn't enough," Lehman stated, emphasizing the need for tools beyond a one-size-fits-all screening paradigm.