Amsterdam Startup Secures Funding to Speed Up Stroke Treatment Development Using AI
Amsterdam, Wednesday 22 July 2026
Amsterdam UMC spin-out inSteps has secured €450,000 to develop an AI platform simulating stroke treatments on virtual patients, cutting testing times from days to minutes.
A Critical Solution for a Leading Cause of Mortality
Stroke remains the second most common cause of death across the European Union, affecting approximately 1.1 million people annually and resulting in hundreds of thousands of fatalities [1]. For patients suffering from the most severe ischemic strokes, medical professionals rely on mechanical thrombectomy instruments to physically retrieve and remove blood clots from the brain [1][2]. However, the research and development pipeline for these highly specialised medical instruments is notoriously expensive, slow, and fraught with clinical uncertainties [1].
The Limitations of Traditional Pre-Clinical Testing
Historically, medical device manufacturers have been heavily dependent on animal testing models to evaluate how new thrombectomy instruments interact with biological tissues [1]. Unfortunately, these animal models are poor predictors of how a device will actually behave within the intricate and highly variable vascular anatomy of a human brain [1]. This predictive gap often leads to unexpected failures during human clinical trials, driving up development costs and delaying the availability of life-saving technologies to patients in critical need [1].
Bridging Finite Element Modelling with Generative AI
To overcome these persistent industry bottlenecks, Amsterdam UMC spin-out inSteps has developed a revolutionary platform that merges Finite Element Modelling (FEM) with generative artificial intelligence [1][2]. This advanced digital-twin platform allows medical device manufacturers to virtually test their thrombectomy designs on thousands of synthetic patients [1][2]. These virtual patient profiles are generated from one of the largest stroke databases in the world, allowing developers to evaluate their designs against highly realistic, clinically diverse anatomical scenarios [1].
Reducing Simulation Times from Days to Minutes
A major technical breakthrough of the inSteps platform lies in its proprietary AI-emulation layer [1][2]. In traditional engineering workflows, simulating the physical interaction between a mechanical device and a blood vessel could take several days of intense computational processing [1][2]. The inSteps AI layer reduces these simulation times from days to mere minutes [1][2]. This drastic acceleration enables manufacturers to identify promising design paths early, iterate rapidly, and bypass the massive financial and temporal costs of physical, trial-and-error experiments [1].
Transitioning from Academic Concept to Commercial Scale
The newly secured €450,000 convertible loan from the Innovatiefonds Noord-Holland in July 2026 marks a crucial phase of growth for the Amsterdam-based startup [1][2]. Led by CEO and co-founder Thomas Kerbusch, the team will use this capital injection to transition their technology from an academic proof-of-concept into an industry-ready commercial product [1][2]. Specifically, the funding will be allocated toward developing advanced modules to simulate synthetic thrombus heterogeneity and the AI-driven emulation of the complete thrombectomy procedure [1].
The Future of Virtual Clinical Trials in MedTech
By providing a scalable, virtual testing ground, inSteps aims to dramatically increase the success rate of clinical trials for stroke devices globally [1]. This regional investment from Innovatiefonds Noord-Holland highlights the robust support system within the Dutch MedTech ecosystem for university spin-offs [1][2]. As digital health technologies continue to mature, virtual clinical trials are poised to become an essential tool in clinical decision-support and medical device regulation [1][GPT].