Dr. Connie Lehman has spent her career looking at mammograms differently. A Harvard Medical School professor and Yale-trained physician-scientist, she has worked for decades at the intersection of radiology and artificial intelligence.
In 2016, she began building an AI model trained on hundreds of thousands of mammograms. That work became Clairity, an FDA-authorized technology that can assess a woman’s five-year breast cancer risk from a single scan. Named to TIME’s 2026 TIME100 Health list and Forbes’ 50 Over 50 Innovators, Dr. Lehman is now moving that technology from academic research into clinical practice across the country.
You spent decades studying mammograms before building Clairity. What were you seeing in that data that convinced you AI could change the way we detect breast cancer risk?
Every day when I was looking at mammograms, I would notice that every woman’s mammogram is like her thumbprint. It is unique to her. In the science and the research, we also found that women’s mammograms were very unique to them, and there were certain patterns on the mammogram that correlated with a woman’s risk of breast cancer.
That happened way before I was in training. In 1967, John Wolfe published the first paper saying patients that develop breast cancer, their mammograms look different years before the diagnosis compared to patients that don’t develop breast cancer. Those observations, we couldn’t really do a lot with them, until we had the power of AI.
The National Comprehensive Cancer Network just updated its 2026 guidelines to include AI-based mammogram risk assessment. How big of a deal is that for women and for the field?
It was such exciting news. I had to pinch myself. We are approaching earlier detection of breast cancer, and the importance of not one size fits all. Screening recommendations are going to be more tailored to the individual woman. We’re going to update our guidelines and include new information so women can benefit now, not in the next generation.
Most people think a mammogram just tells you if you have cancer right now. What else can it actually tell you?
We have been using the mammogram since the late 1960s to see signs of early cancer before anyone would be able to feel it, and we know when we find cancer early, we save lives.
We now know there are so many more data points in every woman’s mammogram. The mammogram now has two powerful contributions to a woman’s individual health. One, do you have cancer now that we need to deal with and treat? Two, what is your risk of developing breast cancer in the future? That second part is the new domain. It’s what we’re so very excited about.
Break down what Clairity can actually do for women
We had a deep learning model. It’s a computer vision model. We gave our model images of mammograms in women who developed breast cancer in five years, labeled as cancer positive, and then we gave it hundreds of thousands of images from women that did not develop breast cancer in five years, and the model learned. Is this a mammogram in a woman likely to develop breast cancer in the next five years or is this a mammogram in a woman who’s very unlikely to develop breast cancer in the next five years?
It turned out that adding in other information about the woman didn’t improve the predictive power of our image-based model. We felt it was going to be the strongest addition we could have to the field, to have an AI image-only five-year breast cancer risk model and put it out into communities so that women had access.
How does Clairity specifically help close the gap for Black women?
Our traditional methods of assessing risk were built largely on European Caucasian women, and we’ve known for decades they do not translate, they do not work for Black women. We knew that, we studied it, we reported it, we talked about it. It’s 2026, we’ve known this for 50 years, and no real change has been made.
When I developed the Clairity Data Partner Consortium, I was like, we will not repeat the sins of the past. We are not going to build a model on European Caucasian women. We are going to form a data consortium with women from Europe, South America, from all over the U.S. Texas to Georgia, Hispanic, Asian, Black women. We wanted that diverse database, and we wanted our clinical validation studies for the FDA authorization to represent the full diversity of women at risk for breast cancer.
What do you want women to understand about how Clairity works and why they can trust it?
The power of data, the power of science, and the power of good research. We study this so carefully. This is within the protected space of healthcare providers. This isn’t going to be something out in the wild that doesn’t have oversight of the FDA, that doesn’t have engagement of our health systems. Women will be supported by healthcare systems. We want to partner with patients and with women to understand what these models can do, what they can’t do, and how much humans are engaged in making sure they’re being used appropriately and in the best interest of every single patient.
For a woman watching this today, what is the one thing you want her to walk away with?
I want every woman to know her risk. We have new ways to assess risk that we never had before, and that information is empowering.
We’re entering a new domain. We’re going to be proactive, not reactive. We’re going to find that risk early, and we’re going to reduce that risk. The idea that with the power of images, we can identify risk early, we can put the right intervention in place to reduce that risk and stop the cancer from developing, that’s the new world.
To know your risk is important, we have new tools available, and we are moving as fast as we can as a small company to build access for more women. 2026 will be a very big year for us. We are now in Boston, so we want women in the New England area to know it’s available near you. We’re launching in Georgia soon, in Colorado. Our website is clairity.com. I wanted AI to be in Clairity.