The traditional path to a new heart valve or stent is paved with animal sacrifice, immense cost, and a frustrating lack of patient diversity. For every device that makes it to a human, years are spent testing on a limited biological sample that often fails to reflect the anatomical and physiological variety of the people who will eventually need it. A spinout from the University of Leeds is betting that a combination of generative AI and physics-based simulation can offer a better, more humane starting point [University of Manchester, February 2024].
Adsilico is building what it calls an in-silico testing platform. The core idea is to use anonymized clinical scans to train AI models that can generate vast, synthetic populations of virtual hearts and blood vessels. These digital twins are then subjected to computational simulations that model how a new implantable device would perform across a spectrum of anatomies, from common to rare [adsilico, retrieved 2024]. The goal is not to replace human clinical trials, but to make the years of pre-clinical work safer, faster, and more equitable [adsilico, retrieved 2024].
The Wedge Into Regulated Device Development
Adsilico's initial applications target some of the most common and high-stakes cardiovascular and neurovascular implants: heart valves, stents, and devices like flow diverters for treating brain aneurysms [The Yorkshire Post, October 2026]. For manufacturers, the platform promises to de-risk early development by identifying potential failure modes in virtual patients long before a prototype is built. This could accelerate R&D cycles and reduce the staggering cost of bringing a new Class III medical device to market, which routinely exceeds $100 million.
The company's technical differentiation hinges on a dual-layer approach. First, a generative AI model creates the virtual patient anatomies. Second, a physics-based simulation engine, built on established computational fluid dynamics and solid mechanics, tests device performance within those virtual bodies. Crucially, the company states it validates its virtual populations against real patient data under clinical oversight and assesses model credibility using the risk-informed principles of the ASME V&V 40 standard, a framework recognized by regulators for assessing computational models in medicine [adsilico, retrieved 2024]. This regulatory-aware foundation is a non-negotiable for any tool hoping to gain traction with device makers who must ultimately seek approval from bodies like the FDA or EMA.
An Academic Engine With Industry Steering
The company is a classic deep-tech spinout, born from years of research at the University of Leeds and the University of Manchester. Its founding team reads like a university department roster, blending expertise in computational imaging, simulation, and medicine.
| Role | Name | Primary Affiliation & Note |
|---|---|---|
| Co-Founder & CEO | Sheena Macpherson | Entrepreneur in Residence at University of Leeds' CISTIB centre [The Org, retrieved 2026] |
| Co-Founder & Head of Virtual Populations | Dr. Nishant Ravikumar | Associated with University of Leeds [The BAE HQ, retrieved 2024] |
| Co-Founder & Head of Simulations | Dr. Zeike Taylor | Associate professor of mechanical engineering, University of Leeds [alphaXiv, retrieved 2026] |
| Founder | Prof. Alejandro Frangi | Bicentennial Turing Chair in Computational Medicine, University of Manchester [University of Manchester, February 2024] |
| Chief Product Officer | Dr. Anthony Holmes | Industry veteran appointed in 2024 [adsilico, retrieved 2024] |
| Chair | Chris Richardson | Veteran in implantable device start-ups, based in Tampa, Florida [Wakefield Express, retrieved 2026] |
The appointment of Chris Richardson as Chair in October 2024 signaled a deliberate push toward commercial and regulatory pragmatism [Clinical Services Journal, October 2024]. His experience navigating the U.S. medical device landscape provides a critical counterbalance to the team's formidable academic prowess.
Funding a Expansion From Heart to Brain
Investors have backed the thesis with a total of £13.5 million (approximately $17 million) in equity funding to date [The Yorkshire Post, October 2026]. Northern Gritstone, an investor focused on university spinouts in the North of England, led the initial £3.5 million seed round in February 2024 [University of Manchester, February 2024]. The recent £10 million Series A in October 2026 was led by PXN Ventures, with participation from existing investors, and is earmarked for expanding the platform's capabilities [The Yorkshire Post, October 2026].
Feb 2024 Seed | 3.5 | M GBP
Oct 2026 Series A | 10 | M GBP
The capital will fuel a planned expansion from the initial cardiac focus into neurovascular devices, specifically targeting the simulation of flow diverters for cerebral aneurysm treatment [The Yorkshire Post, October 2026]. This move into the brain represents a significant technical escalation, given the complexity of cerebral vasculature, but also opens a substantial new market.
The Credibility Hurdle and Competitive Landscape
For all its promise, adsilico's success hinges on a single, formidable challenge: convincing risk-averse medical device companies and, by extension, regulators, that in-silico evidence is credible. The field of computational modeling for regulatory submission is still evolving. While the FDA has shown increasing openness through its Digital Health Center of Excellence and pilot programs, adoption is not yet mainstream.
The company's most significant competition may not be a direct startup rival, but inertia. Device developers have established, albeit costly, pathways to market. Shifting that workflow requires demonstrable proof that virtual testing can predict real-world outcomes and, ultimately, save time and money. Adsilico's answer is to build its methodology on recognized standards like ASME V&V 40 and to pursue rigorous validation partnerships. The next twelve months will be critical for landing flagship design partnerships with device makers that can serve as public case studies.
A New Standard of Care for Pre-Clinical Testing
The patient population at the heart of adsilico's work is broad: anyone who might one day need an implantable cardiovascular or neurovascular device. This includes individuals with aortic valve stenosis, coronary artery disease requiring stents, or brain aneurysms. Today, the standard of care for developing devices for these conditions involves extensive bench testing and animal studies, often in species whose physiology is only an approximate model for human disease. These studies are not only ethically fraught but can be poor predictors of human response, contributing to high failure rates in later-stage clinical trials.
If adsilico's technology is widely adopted, the pre-clinical phase could look profoundly different. Virtual populations could be generated to specifically include anatomical rarities often missed in small animal cohorts, leading to devices better designed for a wider range of human bodies. It could also significantly reduce the number of animals used in research, aligning with a growing global push toward alternatives. The company's bet is that a more predictive, more equitable, and more humane starting point will lead to better devices reaching patients faster. For an industry where time is measured in years and cost in hundreds of millions, that is a proposition worth simulating.
Sources
- [University of Manchester, February 2024] Northern Gritstone and Parkwalk Advisors make £3.5 million seed round investment in ground-breaking computational medicine company adsilico | https://www.manchester.ac.uk/about/news/northern-gritstone-and-parkwalk-advisors-make-35-million-seed-round-investment-in-ground-breaking-computational-medicine-company-adsilico/
- [The Yorkshire Post, October 2026] Adsilico: Yorkshire firm using AI to expand medical device testing raises £10m in funding | https://www.yorkshirepost.co.uk/business/adsilico-yorkshire-firm-using-ai-to-expand-medical-device-testing-raises-ps10m-in-funding-9306127/
- [adsilico, retrieved 2024] adsilico | In silico insights for implantable devices | https://www.adsilico.uk/
- [The BAE HQ, retrieved 2024] Nishant Ravikumar profile | https://www.thebaehq.com/profiles/nishant-ravikumar
- [alphaXiv, retrieved 2026] Zeike Taylor affiliation | https://alphaXiv.org
- [The Org, retrieved 2026] Sheena Macpherson profile | https://theorg.com
- [Clinical Services Journal, October 2024] Chris Richardson appointed as first Chair of adsilico | https://www.clinicalservicesjournal.com
- [Wakefield Express, retrieved 2026] Chris Richardson profile | https://www.wakefieldexpress.co.uk