What if doctors could model how a treatment might work on a patient before ever prescribing it?
That’s the promise of digital twins. Monique Gary, an entrepreneur and breast cancer surgeon researcher with St. Luke’s University known as “Dr. Mo,” believes the technology could bring change to cancer care in the years ahead.
She described a future where AI helps clinicians refine treatment, improve early detection and even personalize what she called “precision wellness.”
“We can refine those drugs and get more precise when we create digital twins.”
Monique Gary, Bexa
“Digital twins is something that is here to stay,” Gary told me. “We can refine those drugs and we can get more precise when we can create these digital twins that say, ‘OK, let’s test out how this is going to respond in women who are 50 years old, who are post-menopausal, who have this stage of cancer.’”
Exciting as that sounds, Gary was just as clear about the caution that has to come with it. “A model,” she said, “is only as good as the population it is validated on.”
I first spoke with Gary while reporting an earlier Technical.ly story about what’s next for AI in healthcare. After that piece published, she reached out to me on LinkedIn because she had more to say. This time, the conversation went deeper on what comes after the current wave of AI tools: not just systems that help with workflow, but advances that could reshape how care is detected, designed and delivered.
Better early detection, thanks to AI
Gary told me AI is already “expanding everything we’re doing in early detection,” especially by making current technologies “smarter, faster, less operator-dependent.”
She serves as chief medical officer at Bexa, a startup that builds a breast tissue analysis device. There, AI is helping with elastography, which looks at the stiffness of breast tissue rather than density to identify possible cancer. She believes that kind of approach could broaden screening options for women who are not well served by one-size-fits-all tools.
“You want to be able to find technologies that can help women who are pregnant, who are young, women who — you know, all walks of life,” Gary told me.
Still, she returned often to something that feels essential in just about every AI conversation: This technology is supposed to help doctors, not replace them.
“We use AI for a lot of those reasons; to help us get smarter and faster and help the clinicians,” Gary said, “and not replace but expand the landscape of early detection.”
Not a shortcut around inclusion
When Gary talked about digital twins, she was speaking as a surgeon, but also as a researcher.
“It is very helpful for us to look at how we can expand both the depth, the breadth, the speed of research through these digital twins,” she told me.
She also warned that the same bias problems that already exist in healthcare research can carry over into AI. When Black and brown patients are underrepresented in trials, the consequences don’t disappear just because the model is digital.
“I don’t want companies to think just because we have some of the molecular … things that we think make up the Black woman now, we don’t need to recruit these Black women into trials,” Gary said. “It’s not a replacement.”
That landed as one of the most important points in our conversation. The future of medicine cannot be built on shortcuts that repeat the exclusions of the past.
Gary also widened the frame beyond diagnosis and treatment. “We focus a lot on precision oncology and precision treatment, but what about precision wellness?” she told me.
For patients in treatment or living with multiple health issues, generic recommendations can miss the mark. “What if you can only take 500 steps?” Gary said, contrasting that with the oft-referenced recommendation to take 10,000 steps per day.
That gets at something bigger than cancer care. It is about whether healthcare innovation actually meets people where they are. And the doctor’s closing point made clear what’s at stake if it doesn’t.
“We have to be at the table,” said Gary, “so we are not on the menu.”
