Expert Views: Could A Digital Twin Save Your Life?

Published 3 months ago
By Forbes Africa | Tiana Cline
new normal Futuristic Technology in medical concept doctor describes patient by using artificial intelligence, machine learning, digital twin, 5g, big data, iot, augmented mixed virtual rality, ar, vr
(Source: Getty Images)

Human digital twins are quickly moving beyond manufacturing and into the medical world – advancing cancer care, speeding up drug development, personalizing clinical trials, and much more.

“You have cancer!” is one of the most terrifying sentences a person will ever hear in his/her lifetime yet it’s something every family is likely to experience. According to the World Health Organization (WHO), cancer is now the leading cause of death globally and the number of new cases is expected to rise by about 70% over the next two decades. When you take into account that there is already an oncology workforce crisis across the continent – South Africa, for example, faces a shortage of radiographers, aging equipment and a decrease in clinical oncologists – providing care to an increasing number of cancer patients is a challenge, more so when the disease is unpredictable and a one-size-fits-all approach doesn’t work. Technological advancements, however, are starting to shift cancer’s unpredictable advantage by merging physical with digital. One way to do this is by creating a digital replica of a physical system or environment – or a digital twin.

Digital twins are already being used in many industries to simulate complex systems and now they’re being adopted in healthcare with management consulting company, Gartner predicting that by 2025, 25% of healthcare delivery organizations will include formalized digital twin initiatives within their digital transformation.


In healthcare, it comes down to the data. What was once trial- and-error is now about feeding millions of data points into complex algorithms for more precise outcomes. With a digital twin, you can simulate an outcome in the digital world before trying it out in the real world, adding in a number of different conditions to see what works best. Human digital twins also make sense because unlike the digital twins used in the automotive industry where each new car model is identical, people are not. And in cancer care, it’s not easy to predict how effective certain medications will be.

“Digital twins are not moving out of the industrial or robotics sector – they’re scaling from the industrial world into the real world,” explains Marc O’Regan, Dell Technologies’ Chief Technology Officer for Europe, Middle East and Africa.

“It gives organizations the opportunity to rigorously test and fail purposely to see what the implications might be, retest again, look for avenues of success and then go back and say – we haven’t even gone into production but we know what the process is going to look like.” Dell has teamed up with the University of Limerick (UL) in Ireland to develop a digital twin for B-cell lymphoma, the sixth most common type of malignant neoplasia in sub-Saharan Africa.

“We’ve built a digital twin for a person with non-Hodgkin lymphoma. We have twinned this particular patient journey through the diagnosis of this cancer and by doing so, we’re able to enrich the data that’s being fed into the twin and for the patient in a positive and impactful way,” says O’Regan.


Dell created an artificial intelligence (AI) platform to produce cancer patient digital twins for better diagnostics. “We will be able to advance our knowledge of how cells go wrong during

cancer development and find new ways to diagnose and treat cancer patients,” adds Professor Paul Murray, UL’s scientific director of the Digital Pathology Unit at the Digital Cancer Research Centre. “It’s not a simulation, it’s a digital twin event associated with all of the componentry on the server supporting that event,” says O’Regan.


—Marc O’Regan


Chloé Audigier is a senior research scientist in the field of AI at Siemens Healthineers. She is currently working on a digital twin of the liver to help identify, simulate and test therapy options for liver cancer – something which could potentially be done before treating a patient. “We are building a general anatomical model that would represent the organ shape, volume and structure. To do that, we are using machine learning algorithms that are trained on the large number of images,” explains Audigier.

“When we have one specific patient, we personalize this model using the organ data of the patient. The model simulates patient-specific physiological function.” Once a digital twin of an organ is created, a virtual simulation of the needed treatment (or intervention) is performed. This gives doctors a clearer and more individualized path to patient care. “We are building computational models that are based on physical law and that we evaluate using outcome data,” adds Audigier.

Creating a treatment strategy for a cancer patient requires an accurate diagnosis. The level of data integration possible with digital twins enables doctors to explore more treatment options and increase precision. Cancer requires planning to optimize tumor control and reduce toxicity – it’s about delivering a high dose of radiation to the tumor while sparing the healthy tissue that surrounds it. “If the digital twin allows us to collect more data, it would be such a gain for radiotherapy,” says Dr Christian Weissenberger, one of Germany’s leading cancer experts. “We can collect the data, we can analyze the data, and we can answer these questions with more validity and more precisely.”

Tomáš Helikar, an associate professor of biochemistry at the University of Nebraska-Lincoln in the United States, compares medical treatment today to ‘playing a game of whack-a-mole’. With his award-winning work to build a digital twin of the human immune system, Helikar’s goal is to be able to control the immune system in a safer and more predictable fashion.


Think of it like a human blueprint that’s always up-to- date and accessible – digital twins built on personalized information mean that doctors would no longer have to base therapeutic regimes on population data but rather adapt to an individual patent’s needs. More detailed knowledge of a patent’s clinical data and physiology can only help to refine decision making processes in the future, allowing specialists to select the best treatment – all it takes is a human, digital twin.