Medical device innovation is laborious because of the high bar for validating that the technology does no harm. Patient recruitment for statistically powered trials stretches timelines and drives validation costs into the tens of millions of dollars. As a result, these evaluations account for approximately 60% of R&D expenditures for complex therapeutic medical devices.
Ethical concerns compound the challenge as some cohorts receive standard care rather than potentially superior treatments. For rare conditions and invasive therapies, assembling adequate study populations is nearly impossible.
To address the clinical burden and enhance R&D, companies are turning to virtual solutions. This involves synthetic data, digital twin models, and AI to speed analysis.
The intelligent solution
So how does this work in practice? Synthetic data is created from real signals such as EEG patterns from people with epilepsy or ECG waveforms from individuals with cardiac conditions. Generative AI then produces variations that represent a broad spectrum of patients and scenarios that are traditionally difficult or unethical to study at scale.
Digital twins integrate the synthetic datasets, anatomical models, and system parameters to create virtual groups. This can enable a cardiac technology provider to test arrhythmia‑detection performance in thousands of virtual patients before enrolling trial participants. This is particularly significant for neurological devices, as this approach can simulate these interactions in orphan diseases.
Enrollment is reduced by generating control groups computationally. This opens up the possibility for synthetic datasets to support algorithm training across scenarios with limited cohorts, which have been almost impossible to statistically validate.
Regulatory environment
A key hurdle for synthetic data to overcome is ensuring regulatory bodies accept simulation-based evidence. The FDA issued guidance on “Assessing the Credibility of Computational Modeling and Simulation in Medical Device Submissions,” which established a framework for in-silico testing. This refers to research, modeling, or experiments performed on computers. By using software, algorithms, and models, researchers can predict safety, efficacy, and molecular interactions faster and at a lower cost, which enhances drug discovery and medical device design.
The FDA’s Digital Health Center of Excellence has highlighted the potential for advanced modeling, AI, and real‑world data to support more efficient evidence generation and development timelines.
Europe has made notable progress, with the EMA recognizing a digital twin as a primary analysis methodology in certain Phase 2 and 3 clinical trials. In 2025, the EMA qualified the first AI‑based tool to support inflammatory liver disease diagnosis, underscoring growing regulatory comfort with machine learning in clinical decision‑making. Then, in December, an EU proposal to streamline MDR and IVDR acknowledged the role of in‑silico evidence in demonstrating device safety and performance.
Virtual validation
Synthetic simulation is gaining traction with the FDA approving over 1,000 AI/ML-enabled medical devices by 2025. While radiology dominates, cardiovascular applications account for a substantial share, with products such as automated cardiac monitoring with AI systems streamlining testing and arrhythmia-detection workflows.
Deploying digital twins has reduced enrollment by up to one‑third, with some designs targeting 35% smaller control arms. In a trial requiring 1,000 patients, shrinking the control group by 25% can cut enrollment time by 4 to 5 months.
Virtual validation is being deployed across applications, spanning automated testing, algorithm training for small cohort diseases, and virtual control groups that curtail physical trial requirements and speed time to market.
Rare diseases come into focus
For individuals with uncommon conditions, these approaches profoundly alter the economics. Prior to this, they were unable to benefit from medtech innovation. Synthetic data removes this barrier, enabling new medical solutions to become commercially viable.
Another area set to benefit is pediatric studies. Virtual patients will provide enough statistical data to expedite drug approvals and access to breakthroughs for children, particularly those with genetic disorders or rare cancers.
For neurological conditions, due to the invasiveness, complexity, and limited number of affected individuals, progress has often been constrained. Therapeutic brain and computer interface devices face particular challenges, as it may take years to acquire enough data. Now, digital twins can simulate these interactions without requiring physical trials. This shortens R&D cycles for treatments that were previously ethically prohibitive.
The economics of synthetic validation
The economic impact is significant. Organizations using synthetic data report up to 70% lower data-acquisition costs and much shorter timelines. This enables faster access to safe, innovative solutions that improve healthcare outcomes.
Another benefit is that the new cost structure helps democratize the medical device industry. With vast clinical trial budgets no longer a prerequisite to market entry, small companies can now compete with established manufacturers. Other factors that will help level the playing field include standards, building anatomical databases, and shared synthetic data methodology. These provide frameworks to facilitate adoption and support regulatory compliance.
The future is synthetic
Healthcare is finally catching up with other industries that have harnessed simulation to drive innovation. However, before this becomes a reality, work must be done around standards, navigating evolving regulations, data quality, and determining the balance of computer-generated and physical testing.
Trial bottlenecks could largely be resolved by the end of this decade as technology makes validation more efficient and comprehensive without compromising safety.
Photo: ipopba, Getty Images
Marie Hattar has more than 20 years of leadership experience spanning the security, routing, switching, telecom, and mobility markets. Before joining Keysight Technologies, Marie served as CMO at Ixia and Check Point Software Technologies. Before that, she served as Vice President at Cisco, where she led the company’s enterprise networking and security portfolio and helped drive its leadership in networking. Marie also worked at Nortel Networks, Alteon WebSystems, and Shasta Networks in senior marketing and CTO positions. Marie holds an MBA from York University and a Bachelor of Engineering degree in Electrical Engineering from the University of Toronto.
This post appears through the MedCity Influencers program. Anyone can publish their perspective on business and innovation in healthcare on MedCity News through MedCity Influencers. Click here to find out how.
