adsilico, a University of Leeds spinout that tests medical implants on AI-generated virtual patients, has raised a £10M Series A led by PXN Ventures.

Heart valves, stents and other implants usually win approval on the back of trials involving a few hundred patients, and recruiting a group that mirrors everyone who will eventually receive the device is slow and expensive. The evidence skews as a result. A JAMA Internal Medicine review of 195 high-risk implant trials found women made up 33% of participants, dropping to 29% in cardiovascular studies. Manufacturers and clinicians often learn how a device behaves in women, older patients or people with unusual anatomy only once it is in wide use.

adsilico's Virtual Chimera technology uses generative AI to create realistic virtual hearts and blood vessels from anonymised clinical scans, then simulates how a device behaves inside each one across a synthetic population that varies by age, sex, ethnicity, body size and disease. Device makers use the results to decide which sizes to build, flag patients who may be at higher risk and sharpen trial recruitment before committing serious money. Regulators are moving the same way: the US Food and Drug Administration has issued guidance on computer modelling as evidence in device submissions, and a UK independent review found that bias in device design can leave some groups with poorer care.

Founded in 2024 by chief executive Sheena Macpherson alongside Nishant Ravikumar and Zeike Taylor, the company has already completed projects with heart valve and neurovascular device makers, and its cardiac work is due to be presented at TCT, a leading structural heart conference. It raised a £3.5M seed round from Northern Gritstone and Parkwalk in 2024, and this round takes total equity funding to £13.5M.

Both seed backers returned, with Northern Gritstone putting in £5M of the total. The money will take adsilico's heart models into brain devices such as flow diverters for aneurysms, where trial numbers are smallest, and grow the team at the university's Nexus hub from 14 to 34. Macpherson said the aim is virtual populations that reflect real-world diversity, making implant testing faster, safer and more equitable.

Sources