A heart valve that fits one patient well can fail another. Yet the trials that approve these devices often enrol only a few hundred people. That mismatch is the problem adsilico, a University of Leeds spinout, wants to fix, and investors have just put £10 million behind it.
The Series A round was led by PXN Ventures, with existing backers Northern Gritstone and Parkwalk also taking part. Founded in 2024 by Sheena Macpherson, Dr Nishant Ravikumar and Dr Zeike Taylor, the company builds software that helps manufacturers test implant performance before clinical trials start.
The problem with small trials
Trials for devices such as heart valves and stents typically involve a few hundred patients. That is rarely enough to represent the range of people who will eventually receive them, whether they differ by age, sex, ethnicity, body size or comorbidities. Recruiting across those groups is also expensive and slow, so many trials simply don’t.
The consequence is uncertainty. A clinician may not know how well a device performs in a patient whose anatomy was barely represented during testing. And the patient may not know either.
How Virtual Chimera works
adsilico’s platform, called Virtual Chimera, uses generative AI to create virtual hearts and blood vessels from anonymised clinical scans. Physics-based simulations then test how a device behaves inside those models. Manufacturers can run the same implant through large, varied virtual populations rather than relying on a small physical cohort.
Companies can use the results to make practical decisions, including:
- Choosing appropriate device sizes
- Identifying patients at higher risk
- Refining clinical trial strategies
The company says the approach adds evidence alongside existing testing methods rather than replacing them. It aims to flag problems earlier, support more targeted studies and potentially reduce reliance on animal testing.
Early traction and next steps
adsilico has already completed projects with heart valve and neurovascular device manufacturers. Its cardiac research is due to be presented at TCT, a meeting focused on structural heart therapies. That’s a useful audience, since interventional cardiologists will be the ones judging whether simulated evidence holds up.
The new money will go toward further development of the virtual heart models and expansion into neurovascular devices, including flow diverters used to treat brain aneurysms. The company also plans to grow its workforce as it moves into other therapeutic areas.
Why it matters
Regulators and manufacturers have been paying closer attention to in silico evidence for several years, partly because of cost and partly because of pressure to reduce animal use. Representation is the sharper argument here. Macpherson, the company’s CEO, put it plainly: patients can miss out on a life-saving device, or face doubt about whether it will work for them, because people like them weren’t represented in design and testing.
Early insight into anatomical diversity is most valuable while a design can still change. Fixing a sizing problem on a computer is far cheaper than discovering it after implantation.
Still, simulation is only as good as the data and validation behind it. The key questions for adsilico are how well its virtual populations match real patients, and how readily regulators will accept the evidence. If those answers hold up, virtual cohorts could become a routine part of device development, and a fairer one.