If you had this type of “twin,” health conditions could be predicted and your digital counterpart could be experimented on to find the best way to treat—or even prevent—disease.
Researchers are interested in developing digital twins that could integrate known human physiology and immunology with an individual patient’s clinical data in real-time, then produce predictions of what would happen during a viral infection, such as COVID-19.
How Digital Twins Could Be Used
James A. Glazier, PhD, professor of Intelligent Systems Engineering at the Indiana University Luddy School of Informatics Computing and Engineering, and Director of the school’s Biocomplexity Institute, tells Verywell that as a concept, digital twins have been around “for 30 or 40 years now, primarily in engineering fields.”
To weave in health information, Sluka says that “the data can then be augmented by basic clinical tests like serum cholesterol, blood sugar, or any other data the physician has already collected for the patient. Up to this point, the data collected is the same as what the clinician has already collected."
For example, jet engines on passenger planes have a computer replica that is continuously predicting what the real engine should be doing and comparing that prediction to the behavior of the real engine. Glazier says that “by looking for deviations from the expected behavior they can predict failure."
Making Predictions, Determining Treatment
If a human patient had a digital twin, it could help doctors predict what the person’s immune reaction would be to viral infections or other medical conditions. Armed with that input, doctors could then run controllable experiments on the twin. Based on the results, which would show the possible outcomes, doctors would be in a better spot to choose the best course of treatment.
Medicine as an Open Loop System
Glazier says that medicine is currently a reactive system or an open loop. “You go in to see the doctor when you get sick, they give you a treatment, and you wait to see what happens," he says.
If the first treatment doesn’t work, your doctor tries something else and waits for a result.
“If we can make those kinds of predictive forecasting software tools, then we’re in a position to begin to design medical interventions that are closed loops, that are preventive, and that are truly personalized,” Glazier says.
How Close Are We to Having Digital Twins?
Glazier says that no one is close to creating a complete digital twin for a human being just yet—primarily because, compared to a jet engine, there is much more complexity and uncertainty in the way that the human body works and reacts.
“One of the biggest holdups that we have at the moment, that COVID has revealed, is that we really don’t understand the human immune system,” Glazier says. “While we can’t do that for the whole body yet, there are increasingly places that we can do it.”
Digital twins are already in use at least on a limited scale, such as monitoring the behavior of one organ or an organ system and then reacting to a situation.
James P. Sluka, PhD, senior scientist with the Biocomplexity Institute, tells Verywell that “there are already a few digital twins in use for specific diseases."
For example, Sluka says that blood insulin monitoring systems like the FreeStyle Libre—insulin pumps that check blood glucose levels and inject insulin as needed—are already useful to some patients with diabetes.
Glazier says that another example of a limited form of a digital twin is an implanted pacemaker that monitors heart rate and corrects an arrhythmia that’s detected.
Digital Twins in the Time of COVID
Glazier and Sluka are coauthors, with Reinhard Laubenbacher of the University of Florida, of a perspective article in Science on using digital twins in viral infections like COVID-19.
The pandemic has required researchers to rapidly recalibrate the computer models that are used by epidemiologists to help public health officials make predictions and create plans to deal with a viral outbreak.
However, we still do not have models to help us predict or explain why different people react differently to the infection (for instance, why a healthy young person dies from COVID while an older adult with an underlying condition survives).
Sluka says that a digital twin could also incorporate a person’s pharmacogenomic data—information from a person’s DNA on how well or poorly they react to certain drugs.
In the future, Sluka says that “complete genetic profiling, whether for prediction of optimum drug therapy or as a more general set of patient-specific data, will be a powerful tool, but in the short term that is not required to build a useable digital twin."
Focusing on Prevention
Once developed, Sluka says that digital twins would “most likely start off simple and then grow in complexity over time.”
In the long term, Sluka says that a digital twin would allow preventive medicine efforts to be tailored to the individual. “For example, at what age and how often should a particular woman receive a pap smear or breast cancer screening? How often should an individual patient get a colonoscopy or chest X-ray?” he says.
“Physicians are already making decisions based on characteristics of individual patients,” Sluka says. “But what is lacking is the ability to rationally make those decisions and to constantly update the decision based on the most current data.”
