skip to main content
Close Icon We use cookies to improve your website experience.  To learn about our use of cookies and how you can manage your cookie settings, please see our Cookie Policy.  By continuing to use the website, you consent to our use of cookies.
Global Search Configuration


Recent explosions in big data gathering and computational learning mean that advances in health care can now become reality. 

One such transformative technology is digital twins, otherwise known as a reflection of the physical world in a digital format that can visualize and contextualize data.

A recent Accenture survey of 399 health care executives found that a quarter of respondents said their company had experimented with digital twins in 2021, while 66% expect their organization’s investment in intelligent digital twins to increase over the next three years.

The application of digital twins could be revolutionary in areas such as personalized medicine, virtual organs, genomics, business processes, operational efficiency, supply chain management and clinical trials.

Drug and device developers have been working toward making clinical trials cheaper and faster for years. Finally, digital twin technology may be the key to unlocking this particularly rusty lock by predicting clinical outcomes from virtual patients.

In Vivo talked to two startups that have developed disruptive digital twin technology that could transform the patient experience forever.

Unlearn Puts Computation At The Heart Of Trials

Charles Fisher started Unlearn.AI with colleagues Aaron Smith and Jon Walsh four years ago. The biophysicist, mathematician and theoretical physicist had met conducting machine learning research at a virtual reality startup in San Francisco and were looking for a new project. The team investigated investment gaps in artificial intelligence and came across clinical trials as an underserved market.

Having secured its first round of venture funding with a developing business plan, the company began two years of working on machine learning technologies. Eventually the team developed a tool that could create patient level simulations of clinical outcomes.

The company did not set out to “solve this problem of clinical trials,” its CEO Fisher told In Vivo. The scientists wanted to solve a challenge in machine learning, and after working through iterations of the product with pharmaceutical partners and advisors, they developed deep learning models that could generate clinical predictions, or digital twins. These digital twins describe what would have happened if a specific patient had received a placebo in a clinical trial. Unlike other approaches that leverage existing data, incorporating digital twins into clinical trials enables smaller, more efficient trials without introducing bias.

The foundational data used to train the disease progression models emanates from historical longitudinal clinical trials, observational data and electronic health record (EHR) data. These data are aggregated through partnerships with pharma and academia.

Charles Fisher

Unlearn is currently focused on Phase II and III studies in neuroscience and immunology because of the large unmet need in these therapeutic areas, and the well-known challenges that occur during these types of trials such as study run time and difficulties in recruitment.

Fisher explained that there is no higher probability of an incorrect result from trials using Unlearn’s digital twin technology, something that is mathematically proven and demonstrated empirically. The Type 1 error rate is unchanged, set to 5% probability of incorrectly rejecting the true null hypothesis.

Unlearn is partnering with leading pharmaceutical companies to accelerate late-stage clinical trials and is actively engaged with the FDA and EMA. Fisher is looking forward to expanding into other indications based on future partnerships.

In Vivo: What does the pharmaceutical industry need to know about your technology? 
Charles Fisher: We want to enable pharma companies to run faster clinical trials, but we want those clinical trials to still produce reliable evidence. We can make clinical trials faster by making them smaller. We can reduce the number of patients in the clinical trial by 25%, that can speed up the time to market by a year, potentially, for some indications.

We use digital twins within the context of smaller randomized controlled trials in a way that enables us to get all the same statistical properties that you want out of a larger clinical trial, that also satisfies existing regulatory guidance. Existing approaches like external control arms can increase efficiency and decrease the size of the trial, but they are not acceptable to regulators in many cases.

The approach that we’ve developed is unique. We can guarantee that it won’t introduce bias and that you get valid results out of clinical trials in a mathematically guaranteed way. This gives us a different position when it comes to regulation, because we don’t need to change today’s regulations from the FDA and EMA, what we do is acceptable.

We’re going through our own qualification procedures now with the EMA and hope to hear back from them within the next couple of months. We’re doing our own work to make sure that we can demonstrate that these technologies are acceptable to regulators.
What feedback have you had from the pharmaceutical industry so far?
There’s lots of interest in the concept of running a clinical trial one year faster. The extra year of revenue is enormous. If the drug works then patients can get that drug a year sooner. The value is clear. We’re shrinking the number of patients who need to receive a placebo. We can’t run trials in which zero patients receive placebo at Phase III because you need to have some randomization — that is the only technology that allows us to be sure that the results are because of the drug. However, we don’t need an equal number in both arms of the clinical trial, so the patient has a much higher probability of getting the active treatment.

There is a lot of uncertainty about AI, how it can be used and what the regulators think about it. Our initial partnerships are with more progressive pharma companies that have initiatives to bring in new technologies.
Do you think the pharmaceutical industry is ready for this kind of disruptive tech? 
I think pharma is totally ready for it. A lot of pharma companies to say they would absolutely love to be our second customer. The regulators are extremely supportive. When we have confirmation from the regulatory direction everyone will adopt these approaches.

We can demonstrate that this is a fundamentally better way of running a clinical trial, a smaller, faster, less expensive clinical trial and we can guarantee that it produces the same level of evidence as larger studies. It is still early days, but we think that within a year or two we will see a wide variety of options in the disease areas we work in.
What opportunities do you see in the market for further digital twin development?
Clinical decision support. A digital twin is the output of a computational model that allows me to ask ‘what if’ questions about a person. Imagine if a physician could access a patient’s digital plan, even remotely, and ask questions about their response to current treatment and envisage how that patient would react to a different medication. There are a lot of challenges to making that reality happen. There are technical challenges between where we are today and what models are required for that. We also need to think about who would pay for that technology.
Do you think this kind of technology can only come from a startup rather than a large pharma? 
From my perspective, those companies [large pharma] should place more value on computation. They’re starting to change a little bit, but it’s slow moving. Company leaders are not computational scientists, they are physicians, biologists, chemists and finance leaders.

We are a computation-first organization. We started to solve problems in computation and AI. I think it’s very different approaching this from a life sciences company with an overlay of AI, whereas Unlearn is AI at its core, and that’s fundamentally different. Computation is going to disrupt a lot of the life sciences industry.


Virtonomy’s V-Patients Simulate Device Impact

Munich-based is the brainchild of Simon Sonntag and Wen-Yang Chu. Sonntag satiated his entrepreneurial desires in 2019 after a decade of academic research and industry experience of advising medical device manufacturers. During this time, he had become heavily involved in the regulatory approval process, especially digital methods such as computational modeling and simulation. He worked on the ISO (International Organization for Standardization) guidelines to help bring in digital elements to the approval process and saw the potential of digital twins and computer simulation, not just for product development but also for product approval.

In the last two years, despite the realities of fundraising during a pandemic, Virtonomy employs more than 10 people, has received seven-digit seed funding led by Dieter von Holtzbrinck Ventures, and has launched its first product, v-Patients.

Simon Sonntag

This digital twin cloud-based technology, offered as Software as a Service (Saas), enables medical device manufacturers – from product concept to post-market surveillance – to perform virtual testing by simulating the device in their target population based on real world evidence data.

Data is gleaned from clinical partners and passes through secure networks. The verification and validation of all steps is essential, Sonntag, Virtonomy’s CEO, told In Vivo, to ensure the credibility and accuracy of the results.

In Vivo: What does the medical device industry need to know about your technology?
Simon Sonntag: Virtonomy can shorten the time-to-market of medical devices by conducting data driven studies on virtual patients. V-Patients is based on an ever-expanding database to reflect anatomical variability, demographic diversity and pathological conditions.

These computer simulations on real clinical data have the potential to reduce the cost, and the time to market, by up to 50%. And above all, it also reduces the risk of failure. There are a lot of iteration loops that can be performed early on during the product development stage, so once you then enter the clinical trial domain you can be more confident in the safety and performance of your product. You can pressure test it in ways that you would never be able to if you used conventional approaches. You can elevate and extrapolate the tests to thousands, possibly even millions, of simulated scenarios to represent the variability of the whole patient population.

Diversity in clinical trials is still a major problem. Simulated studies can conduct trials on not only a larger scale but a more representative sample of patients.
What can be done to speed up adoption of digital twin technologies?
The regulators must understand the evidence because it is new to them. Regulatory uncertainty is still the major concern for the medical devices industry. A lot has happened within the last year, especially with the FDA. It has taken some big steps toward accepting digital evidence, but also on providing guidance documents such as the guidance on reporting of digital evidence, and it has been working with ASME on verification and validation standards.
What opportunities do you see in the market for further digital twin development?
There is a market for treatment elevation and personalized medicine. Digital twins are used to represent the individual so, once a clinical decision is made, they could also be used to predict medical outcomes. It is like travelling into the future.
Is a startup better placed to drive this kind of innovation in the market rather than a large company?
Companies like Virtonomy have huge potential as it’s a new market. We can be very innovative and fast moving, which is unlike large medtech or pharma companies. But, of course, smaller companies need financial power. Investors are increasingly seeing the potential within this field and supporting startups. If you can couple an innovative nature with the financial power of investors there is a clear advantage to smaller companies such as ours. 

Read also


Next steps

Whether you’re a small biotech start-up, research firm, generic manufacturer or a global pharmaceutical giant, you need focused, independent insight and opinion on market developments.

Our team is ready to hear from you for a particular request or area of interest. Please do not hesitate to reach out and discuss.

Contact us for product technical and account support.

  • US Toll-Free   : +1 888 670 8900 
  • US Toll             : +1 212-600-3520
  • UK & Europe : +44 (0) 208 052 0700

Have an immediate and specific information need?

Browse and buy from 1000s of analysis and research reports now: