Digital Twins: Accelerating Medical Research and Device Development

This article is the initial installment in a six-part series exploring the influence of artificial intelligence on medical research and treatments. A simulated heart, a “digital twin,” beats and moves akin to a human organ, yet lacks blood flow and exists outside a human body. This computer-generated model serves to evaluate implantable cardiovascular devices, such as stents and prosthetic valves, ensuring their safety before eventual use in real patients. Adsilico, the creators of this virtual heart, have advanced beyond a single precise model. Utilizing artificial intelligence and extensive data, they have developed numerous distinct hearts. These AI-generated synthetic hearts can incorporate not only biological characteristics like weight, age, gender, and blood pressure, but also specific health conditions and ethnic backgrounds. Since these diverse characteristics are often underrepresented in conventional clinical data, digital twin hearts enable device manufacturers to conduct trials across a broader range of populations than would be possible with human trials or trials relying solely on digital twins without AI. Adsilico chief executive Sheena Macpherson states, “This allows us to capture the full diversity of patient anatomies and physiological responses, which is not possible using conventional methods. This use of AI to enhance device testing leads to the development of devices that are more inclusive and safer.” In 2018, an investigation by the International Consortium of Investigative Journalists disclosed that medical devices were responsible for 83,000 deaths and over 1.7 million injuries. Ms. Macpherson expresses optimism that AI-powered digital twins can reduce these figures. Ms. Macpherson, based in Northumberland, explains, “To really make these devices safer, you need to test them more thoroughly, and it isn’t feasible to do that in a clinical trial environment due to the expense of it. So you want to be able to use the computer-generated version, to make sure that whatever you’re doing, you’ve tested it as thoroughly as possible before you test it on a human. Even a fraction of those deaths – and the associated lawsuits – could have been avoided with more thorough testing. You can also get more detailed results. You could take the same [virtual] heart and you could test under low or high blood pressure, or against different disease progression, to see whether that affects the device in any way.” Ms. Macpherson further notes, “[Virtual] testing gives medical device manufacturers many more insights. It also means that we can test in other sub patient groups, not just white men which clinical trials have traditionally been based on.” Adsilico’s AI models are trained using a combination of cardiovascular data and information from actual MRI and CT scans, including medical imaging provided by consenting patients. This data leverages detailed anatomical structures of the heart to generate precise digital representations of how medical devices will interact with various patient anatomies. Adsilico’s trials involve creating a digital twin of the device under examination, which is then placed into a virtual heart within an AI-generated simulation. This entire process occurs within a computer, allowing the test to be replicated across thousands of other hearts—all AI-simulated versions of a real human heart. In contrast, human and animal trials typically include only hundreds of participants. A significant motivation for drug and device manufacturers to augment clinical trials with AI digital twins is the reduction in development time, which also leads to substantial cost savings. For instance, drug manufacturer Sanofi aims to decrease its testing period by 20% while simultaneously improving its success rate. The company employs digital twin technology in its immunology, oncology, and rare disease specialties. Sanofi utilizes biological data from real individuals to create AI-based simulated patients—distinct from actual clones of specific people—which can be distributed among the control and placebo groups within a trial. Sanofi’s AI programs then generate computer models of the drug to be tested, synthesizing properties such as how the drug would be absorbed throughout the body. This allows for testing on the AI patients, with the program also predicting their reactions, thereby mimicking the actual trial process. Matt Truppo, Sanofi’s global head of research platforms and computational research and development, states, “With a 90% failure rate across the industry of new drugs during clinical development, an increase of just 10% in our success rate by using technologies like digital twins could result in $100m in savings, given the high cost of running late phase clinical trials.” Mr. Truppo, located in Boston, US, adds that the outcomes thus far have been promising. He notes, “There is still a lot to do. Many of the diseases we are now trying to impact are highly complex. This is where tools like AI come in. Powering the next generation of digital twins with accurate AI models of complex human biology is the next frontier.” However, digital twins may have limitations, according to Charlie Paterson, an associate partner at PA Consulting and a former NHS service manager. He highlights that the effectiveness of these twins is contingent on the quality of the data they are trained on. Paterson states, “[Due to] aged data collection methods, and low representation of marginalised populations, we could end up in a position where we could still be introducing some of those biases when we’re programming virtual recreations of individuals.” Sanofi acknowledges and is addressing the challenge of training its AI with limited legacy data. To bridge gaps in its internal data sets—comprising millions of data points from thousands of patients participating in its trials annually—the company acquires data from third parties, including electronic health records and biobanks. At Adsilico, Ms. Macpherson expresses optimism that AI digital twin technology will eventually eliminate animal testing from clinical trials, a practice currently considered vital for drug and device testing. She remarks, “A virtual model of our hearts is still closer to a human heart than that of a dog, cow, sheep, or pig, which tends to be what they use for implantable device studies.”