Beth Israel starts using AI-driven tool to spot breast cancer risk
Add Axios as your preferred source to
see more of our stories on Google.
Illustration: Allie Carl/Axios
Beth Israel Deaconess Medical Center on Friday rolled out an AI-driven technology that detects the risk of developing breast cancer in the next five years.
Why it matters: The Clairity Breast platform, which obtained FDA de novo authorization last year, can detect breast cancer risk for a broader swath of people, says Connie Lehman, Clairity's founder and CEO.
How it works: Clairity uses an AI risk model that was trained on over 421,000 images of mammograms belonging to both patients who did and did not eventually develop cancer.
- That taught the model to detect subtle patterns in breast tissue that later develop cancer — patterns that doctors can't see themselves.
- The model was trained on images of patients of various ages, races and ethnicities, Lehman tells Axios.
What they're saying: "We're actually showing an application of AI for good," says Lehman, who is also a breast imaging specialist at Massachusetts General Hospital and professor at Harvard Medical School.
- "We're addressing some of the biggest problems in breast cancer prevention."
Threat level: While early detection and treatment of breast cancer has improved, especially for those with a genetic risk, the disease is on the rise.
- Breast cancer incident rates are growing faster among women under 50.
- Women under 40 are now more likely to develop more aggressive forms of breast cancer.
Lehman says the platform captures breast cancer risk that might otherwise go unnoticed because it's not genetic.
Yes, but: The technology isn't covered by any insurance yet.
- The test costs roughly a couple hundred dollars, Lehman says.
What we're watching: Clairity is working on getting it covered through the Centers for Medicare & Medicaid Services and other insurers.
- "I would like to see that everyone woman who gets a screening mammogram can also receive her Clairity Breast score," Lehman says.