A doppelgänger that could save your life? Only at Duke.
What if you could know the outcome of your heart surgery before the operation begins? Understand how your body will react to chemotherapy before you seek treatment? Even predict whether you will develop a life-threatening disease 10 years from now?
Those are exactly the answers biomedical researcher Amanda Randles seeks to answer through a new “digital twin” generated from software she created here at Duke.
A digital twin is an exact virtual replica of something that exists in real life, accurate right down to the cellular level. The twin can be used to predict disease, prevent life-threatening conditions and personalize treatments, providing patients with safe, non-invasive health care right at their fingertips.
“Our goal is to make things as patient specific as possible, and to give doctors real-time feedback,” Randles says.
The software Randles and her team created at Duke is called HARVEY, named after William Harvey, a 17th-century surgeon who is credited with first describing the circulatory system. The software takes 3-D images of a patient’s blood vessels and then simulates and forecasts precisely how the blood will run through the body. To do this, the software looks at the unique size and shape of a person’s blood vessels and the dynamics of blood flow. This is a dramatically different approach than how physicians must currently check for blockages—by placing a stent into a patient’s blood vessels.
Randles hopes that the ‘digital twin’ will allow physicians to take a less invasive approach. Doctors, surgeons, even physical therapists will be able to “try out” treatments on a digital twin in a consequence-free world.
“Everyone can appreciate it would be better to have something non-invasive rather than having to have an implantable device that’s going to track what you’re doing or have to go back to the doctor’s office and spend hours there,” Randles says.
The amount of data points involved in this kind of complex biological modeling are staggering. Indeed, Randles’ biggest research breakthrough have been in creating entirely new ways to deal with such a vast set of numbers. Her innovations in high-performance computing have allowed her to develop the new computational tools necessary to handle the numbers involved in perfectly replicating a person’s unique circulatory system in real time.
Randles has disclosed multiple inventions based on this research to the Duke’s Office for Translation & Commercialization, which now has multiple issued patents and is helping Randles bring her technologies out of the lab and into the real world.
Her lab is collaborating with Duke Health to enroll patients into clinical trials with carotid disease who have had surgery. They are given a Fitbit and are being tracked over several months to see if physicians can predict, using these 3D models, whether the patient is at risk for another blockage of the carotid artery.
“We’ve shown mathematically that it is possible to remotely monitor using these wearable devices. But to translate that into something that’s going to have an impact on people, we need to do the clinical studies, and we need to do more verification and validation before it can be used by the general public. And we need NIH funds for any of that,” Randles says.
Her lab has mostly been funded by the National Institutes for Health. Without future funding, her work may not be able to move forward. For patients, that might mean delays in diagnosis and treatment. In addition to her work on cardiovascular disease, Randles is using the technology to look at cancer cells, what drives disease development, and what leads to metastasis.
“If we want to have any chance of addressing any of these pressing concerns, we need to understand why they’re happening, where they’re happening, and how to best treat them,” she says.
