As the saying goes, everyone has a twin. And in a few years, that might be true — at least digitally.
In an ongoing effort to use artificial intelligence and machine learning to improve personalized medicine, some medical centers are employing programs that enable them to take all of a patient’s data to create that person’s so-called “digital twin.”
In medicine, this is achieved by combining data about a person’s genes, proteins, cells and personal data to create a virtual model of their organs or whole bodies, according to The Guardian. Consulting firm Gartner, Inc., predicted in 2022 that 25% of health care delivery organizations would utilize digital twin projects by 2025, according to Forbes.
Uses of digital twins
The technology behind digital twins was first used by the space program in the 1960s to model spacecraft development and simulate the moon landing, according to an article in HealthLeaders Media. It has since been used to design buildings, roadways, theme parks, planned communities and more.
Digital twins entered health care about a decade ago with the Living Heart Project, a collaboration between industry and the FDA to get cardiovascular researchers, medical device developers, cardiologists and more together to develop highly accurate personalized 3D human heart models. The idea was to use those models to design cardiac technology and plan surgeries, treatments and other procedures.
Now, digital twin models are being created for additional body parts such as the brain, liver and lungs. Boston Children’s Hospital is partnering with industry on applications of the technology for pediatric care, according to HealthLeaders Media, while Johns Hopkins in Baltimore is exploring the potential for digital twins for remote surgery and treatment.
In 2022, the University of Miami announced it would use federal funds to create a digital twin of a patient using data collected from wearables and sensors. Called the MLBox, the twin could be used to test new treatments before they’re tried on an actual patient.
The University of Pittsburgh Medical Center in early 2023 also said it was testing digital twins, according to Becker’s Health IT.
“We call it a ‘digital twin’ because the technology allows clinicians to create a model of the patient with all the same demographics, health conditions and medications,” said Chris Carmody, the hospital’s chief technology officer, in an April 2023 university article. “With this model, the care team would be able to adjust medications or other treatments and see potential outcomes or side effects. It would allow clinicians to test new treatment plans before enacting [them] so they can determine the best way forward.”
At a November digital twins conference in London, as reported by The Guardian, presenters discussed various applications for digital twin technology, including:
- Developing personalized heart models to help surgeons plan surgery for patients with irregular heartbeats.
- Building digital replicas of cancer patients’ tumors to see how they respond to drugs.
- Building digital twins of pregnant people to help develop drugs for conditions such as preeclampsia and to better understand physiological processes behind pregnancy and labor.
In another use of the technology, Thomas Jefferson University Hospital in Philadelphia partnered with artificial intelligence company BigBear.ai in October to build a digital twin of the hospital itself. This was done to test improvements in central resources and bed management, nurse staffing and more, in part to see how the hospital would handle stress during periods of change.
Story ideas
Journalists can find interesting stories by looking into any of the potential use cases for digital twins outlined here. More can surely be found by searching through medical literature or asking around various academic medical centers or informatics groups such as the American Medical Informatics Association.
Ask:
- What’s the status of these projects?
- What are they finding?
- What are the advantages?
- How likely are the digital twins’ responses to various stimuli to match what will happen with a real person?
There’s also a potential patient angle, if there are patients who have consented to having their information used to create a twin. One example might be the case of someone with a rare disease or unusual condition. Here are some questions you could ask:
- Why did they agree to share their data?
- Are they excited about this possibility?
- What do they hope will come from studying this?
And, there’s a business angle as well: How much are these programs costing versus how much could they save? This is especially true of hospitals that are partnering with third-party technology companies.
Resources:
- How digital twins may enable personalized health treatment – article from The Guardian
- Digital twin tech is set to reshape healthcare dx – article from HealthLeaders Media
- Researchers using digital health tools to design a ‘digital twin’ for patients – article from Healthleaders Media
- How UPMC is integrating ChatGPT, ‘digital twinning’ into patient care – article from Becker’s Health IT
- How Thomas Jefferson is using AI to create its ‘digital twin’ – article from Becker’s Health IT
- How digital twins can accelerate health care transformation – article from Forbes.com
- Digital twins: Tailoring medicine insilico – agenda from the Royal Society of Medicine’s recent digital twins conference in London
