How Virtual Modeling and AI Are Enabling MedTech In-silico Clinical Trials

The field of in-silico clinical trials is at the forefront of industry innovation. And advancements in this area—from computational modeling and simulation to generative AI and synthetic data—have the potential to drastically impact medical device development.

Steve Levine, Sr. Director of Human Modeling at Dassault Systèmes and Afrah Shafquat, Sr. Data Scientist II at Medidata joined Device Talks Weekly to detail what MedTech companies can expect from the FDA’s new playbook for digital simulations and in-silico clinical trials. Here’s what you need to know.

What’s the Origin of the In-silico Clinical Trials Playbook?

The in-silico clinical trials playbook is the end-product of a decade-long industry collaboration led by Dassault Systèmes and the FDA, known as ENRICHMENT In-Silico Trial Program. This project is highlighted in the playbook and supports the FDA’s 21st Century Cures Act by using virtual human modeling and generative AI to develop in-silico clinical trial approaches that improve the efficiency of studies.

The FDA has recognized the public health benefits offered by modeling and simulation, as well as the opportunities that in-silico clinical trials present to advance medical products more efficiently. And much progress was made between Dassault Systèmes and the FDA by using computational models (also known as virtual twins) as fully functional replications of human anatomy, electrophysiology, hemodynamics, and more—in this case, the living heart project.

How Are Virtual Modeling and Simulants Advancing In-silico Clinical Trials?

The living heart is a virtual twin—a complete replication of the human heart. This breakthrough can reproduce any patient, disease, or population with a fully functional, virtual model. And with the help of AI methodologies, this reference model can move through the full lifecycle of a patient—from reproducing the disease condition, to virtually treating it and reviewing the outcome.

The virtual twin uses physics-based models from a patient’s measurements to capture the necessary properties. But these models can take weeks to produce the required data. Generative AI can accelerate the physics-based parameter model data and generate synthetic data for a patient population in less time.

How Can In-silico Clinical Trials Benefit the MedTech Industry?

Generative AI and multi-scale modeling can be used together to speed up product development augmenting clinical studies with in-silico data to accelerate the development process, lower costs and improving patient safety.

Additionally, there’s a wealth of knowledge and information learned by medical device sponsors during clinical studies that don’t make it to the FDA and reviewers, slowing down the review process. The ENRICHMENT project detailed in the in-silico clinical trials playbook shows how computational modeling of an entire trial would let reviewers understand exactly what will happen before the trial is run.

When Will the In-silico Clinical Trials Playbook Be Released?

The FDA, assisted by Dassault Systèmes and other industry collaborators, released final guidance on how to use computational models in the regulatory process. The in-silico clinical trials playbook is an advanced use of these guidelines and is expected to be released in a peer-reviewed journal in 2024.

The playbook will describe how to model not only a medical device or patient, but also an entire population, to accelerate successful treatments.

Steve Levine and Afrah Shafquat discuss the in-silico clinical trials playbook.

Supporting Medical Device Clinical Trials

MedTech leaders need a trusted partner to bring products to market faster, amplify top-line growth, and succeed. Learn how companies achieve clinical, regulatory, and long-term success with Medidata.