Virtual Twins: Their Roles In Healthcare, Drug Discovery, And Pandemic Response

By Allison Proffitt

July 9, 2020 | In a world that has shifted to virtual everything—work, recreation, connection—it’s no wonder that virtual or digital twins are getting more attention. It’s certainly a favorite term of marketers: we are creating digital twins for everything from people to businesses, instruments to labs.

Steve Levine, the senior director of virtual human modeling at Dassault Systèmes, has been immersed in this world for years and he acknowledges that the terminology, “can get a bit complex.” But he also believes that right now—in the midst of a pandemic that is impacting everything from engineering to healthcare—there are great opportunities for virtual models.

“Each day I’m realizing new ways that this technology can really make a difference—especially given these circumstances. It’s top of mind for so many people: how can we be so advanced but feel so helpless.”

At its core, a virtual twin is a model and Dassault Systèmes has a long history with modeling. In 1989, Dassault Systèmes created a digital twin of a commercial jet, a Boeing 777, Levine explains. “The concept of modeling an entire jet on a computer was unheard of 40 years ago. We couldn’t conceive of the computational power necessary to do that; now it’s done routinely.”

The human body was the next logical step, Levine said. “Our aspirations are to do the same thing with the human body. At each stage with the level of understanding, computing power, resources, etc. available to us, we’ll represent what we know and systematically put it together.”

In 2014, Dassault Systèmes unveiled the Living Heart Project, the first realistic 3D simulation model of a human heart. Levine is the founder and director of the Living Heart Project and he says it’s the first deep investigation into virtual twins. In 2016, the model won a Bio-IT World Best of Show award.

A virtual twin is not unlike what doctors do in their heads, Levine said. “That’s why they spend many years trying to understand their part of the human physiology: so when they get a piece of data, they can apply it to their mental model of you and then know what to do next.”

But our brains can only process so much information, and today there are more data inputs than ever. Not only symptoms and lab values but data from sensors like heart rate monitors, exercise trackers, plus environmental inputs on our habits, where we live, and other social determinants of health. That where building a virtual model helps.

Building Model Complexity

Levine says that to sufficiently understand the human body, we need to build realistic models of humans—a goal of the company’s. But a realistic model doesn’t necessarily need to be an exhaustive model.

“The key to the value of a virtual twin is not the fidelity of the model, it’s the application of the model,” Levine explained. “If the model is sufficient to take the inputs of data that you’re getting and… make predictions, then the model is effective and useful as a virtual twin.”

That is a broad definition—one of the reasons Levine says the terminology, “can get a bit complex.” But he emphasizes that simple models can be as effective as really complex ones—"if you’re modeling the phenomenon correctly.” The goal is not to reproduce nature, but to only add enough complexity necessary to get the answer.

And that level of complexity will change depending on the question you are asking. “Although we have modeled the human heart down to the cellular level, you wouldn’t use that in a day-to-day monitoring way,” Levine pointed out. “It’s too complex.”

Levine considers the model of the full human to be the center of the healthcare modeling continuum: at one extreme are cellular models, at the other are population-level models. “If you’re modeling cellular behavior, it’s very decoupled from the world, yet it’s part of the virtual twin,” he said. Likewise, “a population model is a very important virtual twin of society.”

Security and Regulatory Receptiveness

While Levine is confident that Dassault Systèmes will achieve a complete human model, that doesn’t keep him from singing the praises of single system models. In fact, he argued, models of independent systems alleviate some of the privacy concerns of more complex human models.

“I don’t think you have to replicate a patient in their entirety to be using some of their personal data as a virtual patient,” Levine said. “In particular, many of these drugs and things like that don’t operate on the whole body. They operate on one system. If you can model that system, you have the important part of that person for that drug.”

Dassault Systèmes is running an in silico clinical trial experiment with FDA now to explore how this might impact drug and device development. The team is using the Living Heart Project to create a cohort of virtual patients to test a mock artificial heart valve and submit the data to FDA. Two teams of regulators—one blinded to experiment, one not—will review the dataset. The goal is to see if FDA would be confident taking the virtual patients as a replacement or augmentation to real patients.

Levine hopes this experiment will serve as a playbook for future in silico trials. It’s a promising way to conduct research, he believes, and Dassault Systèmes will release all of the data from this experiment to the industry to inform next steps.

Virtual vs Hypothetical

In the current FDA experiment, the patient model is only of the heart, with little to no identifying characteristics, but it is still based on real patient data. This raises important societal questions about how we, as a society, manage virtual data, Levine pointed out. What if a device company wanted to save that model and use it to test the next iteration of its heart valve? And the next? And the next? What consent would be required?

There is an alternative, he pointed out. Instead of being based on real people, virtual patients could be complete hypotheticals, computer-generated to represent the range of human variables.

In the current study, FDA wants real patient data to validate the model. Their belief is, Levine explained, that to show that a device or compound works in a real person, you have to base the data on a real person. However, all clinical trial cohorts are really just a model of a population, Levine argued. “If our goal is to get to the population, do we really need to go through the individual to get there?”

Levine doesn’t think so. “We should not limit ourselves by how the real world limits us. We can’t create a person that represents more than that person, but we can create a model that represents more than one person. Why not take advantage of that? But we have to know enough to be able to do that and be confident in it.”

Virtual or in silico cohorts may also be a tool for introducing much-needed diversity into clinical research. To start, we still need to identify diversity from real patients, but, “we’re far more alike than we are different,” Levine said. “We start from the same, but we need to make sure we make those adaptations to take on the full diversity,” he explained. “But the advantage of doing it with virtual patients is you don’t have to do it every time. Once you understand the diversity, you can build that into the [future, virtual] patient population.”

Community Models

In building the Living Heart Model, Dassault Systèmes wanted to not only build a complex model of the heart, but to unite the cardiac research community—a challenge far greater, in many ways, than the technical ones. “The knowledge that we have collected as a society is incredible, but because it’s divided into so many different places, it’s like finding the information before we had Google. All of that information was out there, but because there was no way to bring it all together, it wasn’t powerful,” Levine explained. “The Living Heart Model was an attempt to bring those minds and databases together because they were all referencing the same foundation.”

In addition to the Living Heart Model, Dassault Systèmes is working on other systems as well, working toward a full human model. “I’ve gotten a lot of interest in the lungs, it turns out, in the last few months!” Levine said wryly.

In addition, a brain model is in the works and musculoskeletal systems have been developed for knee and hip implants, feet, shoulder and spine issues. The human modeling team is working reverse engineering the musculoskeletal systems so that when a patient has an abnormal finding in a grip test or a walking test, the model can predict precisely what is going on at the muscular level and develop biomarkers that a drug company can use to refine development.

It could be an extremely useful advance, Levine said. For example, knowing that walking speed increased by 2% for some patients isn’t very actionable. But knowing that one set of muscles were strengthened while others were unaffected—“That kind of resolution is fantastic for drug development.”

None of these other virtual systems has recruited the same extensive community as the Living Heart Project yet, Levine explains, but places much of that blame on the pandemic and our current altered reality. He does count himself fortunate that Dassault Systèmes’ reputation from the Living Heart Project has attracted interested partners.

Pandemic Models

Levine believes there are many opportunities for virtual twins to play roles in the current pandemic—both in treating COVID-19 and in modeling our healthcare system. “One of the aspects that I’ve learned, and really try to highlight to people, is not to think of a virtual twin as a standalone,” Levine said. “It’s a tool.”

As the real world gathers more information—like we’re doing with COVID-19, Levine pointed out—models can be developed, updated, and refined. “The goal of the virtual model is to allow you to get that real-world feedback, so that you can understand that information in more depth for a complex system like the human body,” he said.

“But you know, healthcare is not just about the person. It’s about the system. And part of what we’re learning in this pandemic is that the system can get overloaded; it can be suboptimal. Understanding how the system operates on individuals and how that connects through the supply chain is also a big part of the healthcare system,” Levine explained. Manufacturers spend a great deal of time modeling and refining their factories and processes, he pointed out. Healthcare and medical supply chains should follow suit. The end result may not be quite as flashy as modeling a beating heart, but Levine believes it is certainly as important to health.

“The most impactful part of this technology is that healthcare in general is so complex. It involves so many different disciplines that speak different languages, have different ways of thinking, and understanding,” he said. “And they all come together when you actually create a virtual twin.”