Imagine walking into a clinic where your doctor doesn’t just study your reports but turns to a virtual version of you—one that can forecast how your body might respond to a drug, anticipate risks before symptoms appear and test treatments in seconds. This is no longer the stuff of science fiction, but an emerging reality of medical digital twins.
The idea of digital twins began in the early 2000s, when Michael Grieves at the University of Michigan described a digital model of a physical system that could be used to monitor and improve it. NASA was among the first to apply this idea during the Apollo missions, using simulations to diagnose problems without putting astronauts at risk.
As computing power, cloud technology, sensors and AI advanced, digital twins spread across industries such as aerospace and manufacturing. By the mid-2010s, they were being used in medical science.
Unlike a digital twin of a physical object, a medical digital twin goes beyond basic 3D modelling. It is a dynamic, data-linked computational representation of an individual patient that updates over time. These models are designed to predict future outcomes and enable more precise, personalised care.
According to the recent AI Index Report from Stanford, research in this area has grown rapidly, with publication counts rising from near zero in 2015 to 372 in 2025. Patent filings in health care digital twins show a similar trend, increasing from 30 in 2016 to 4,926 in 2025.
Historical and real-time data drive the creation and evolution of a digital twin. This includes anatomical structure, electrical and chemical activity, genetics, environmental factors and continuous inputs from wearables or implants. The model evolves in near real-time, staying aligned with the physical patient.
Clinical trials incorporating digital twin elements accelerated in 2025, particularly in oncology and diabetes. A pilot trial in prostate cancer using adaptive therapy reported significantly improved survival. In diabetes, a randomised controlled trial of Twin Health’s Whole Body Digital Twin platform found that 71 per cent of participants achieved an HbA1c below 6.5 per cent within 12 months, while safely reducing their use of blood sugar-lowering medications.
A key advantage is the ability to predict how conditions such as cardiovascular disease, cancer recurrence or chronic illnesses like diabetes may evolve, with or without intervention. Digital twins can also enable early detection of risks, such as arrhythmias or metabolic changes.
More important, digital twins are expected to become the new norm in testing drugs, dosages and therapies, predicting side effects and efficacy before real-world use. Surgeons may also practise procedures on these models, such as stent placement, ablation or tumour removal, to identify optimal approaches, reduce operation time and minimise complications.
However, challenges remain. These include concerns around data privacy and security, high computational demands, questions of model accuracy and validation, integration with existing health care systems and ethical issues related to predictive insights.