2025 Replays

Dassault Systèmes' Science Week 2025 started with "Science in the Age of Experience: Science UNIV+RSES", gathering worldwide Scientific Communities to leverage the power of virtual universes and platforms

As Patrick Johnson, Executive Vice President of Corporate Research and Sciences at Dassault Systèmes, explains in his welcoming remark, we are now in the Age of the Generative Economy, where the Age of Experiences and sustainability imperatives converge. We believe more than ever that the best way to build a better future for all is through Science, and with Scientific UNIV+RSES, we turn knowledge into impact, reshaping the world through innovation.

Convinced that virtual worlds can improve science practices and outcomes in many strategic domains, such as engineering, material science, medicine, biology, data science or natural sciences, Dassault Systèmes intends to catalyze how sciences open our eyes and make the invisible visible. Virtual worlds allow us to explore, understand, imagine, create, test and elevate our understanding and means to act to resolve our most significant challenges in the real world.

Jos Benschop, Executive Vice President Technology at ASML, outlines how ASML integrates artificial intelligence across the lithography value chain to enable the next generation of semiconductor innovation. As society moves from “chips everywhere” to “AI everywhere,” the unprecedented adoption of generative AI drives explosive demand for advanced computing hardware. Meeting this demand requires continual progress in lithographic scaling—achieved through innovations in wavelength reduction, numerical aperture, system precision, and high-performance EUV technology.

The talk highlights how AI is increasingly essential both as an enabler and a user of advanced lithography. Applications span design optimization through surrogate modeling, faster and more accurate optical proximity correction, physics-informed control of EUV source parameters, and digital-twin–based scanner optimization. These AI-driven approaches significantly accelerate development cycles, improve accuracy, and reduce the reliance on manual engineering iterations. The presentation also emphasizes the extreme engineering challenges behind EUV lithography—requiring picometer-level optics precision, high-power laser systems, and sophisticated motion stages—and the collaborative ecosystem necessary to achieve such breakthroughs. Together, these advances illustrate how AI and holistic lithography jointly drive the performance, efficiency, and scalability of future semiconductor technologies.

Patrick Nebout, Chief Technology Officer at Yanfeng Technology, explores how Yanfeng is redefining the future automotive cabin at the intersection of user experience, sustainability, and engineering complexity. As the automotive landscape evolves beyond the traditional CASE framework, accelerated development cycles, personalized in-cabin experiences, and the growing integration of AI are reshaping expectations—particularly in China, where advanced Level 2 autonomy and intelligent interiors are becoming mainstream. Yanfeng presents its strategy as a global cabin supplier, supported by deep design, engineering, material, and manufacturing expertise, and strengthened by a broad ecosystem of OEMs, suppliers, and research partners.

The talk highlights emerging cabin concepts for Gen Z users, emphasizing flexible interior layouts, immersive digital experiences, and seamless transitions between multiple usage scenarios. A special focus is given to next-generation seating systems, where multiphysics modeling, human-factor analysis, and comfort engineering converge to deliver innovations such as the Hover Seat and zero-gravity seating with enhanced safety features. The presentation also illustrates how system engineering, model-based validation, and virtual-physical integration enable teams to collaboratively manage cabin complexity, optimize performance, and reduce development time. Through its XiM25 Experience-in-Motion concept, Yanfeng showcases a holistic vision of future mobility that blends aesthetic design, sustainable materials, user-centric functionality, and advanced engineering into a unified cabin ecosystem.

From traditional to intelligent to generative, learn how Digital Twin Consortium members are at the forefront of the evolution of Digital Twins. This presentation by Dan Isaacs, Digital Twin Consortium CTO, will provide an overview that traces the primary attributes of Digital Engineering over the lifecycle of the Composable Digital Twin through real-world use cases. The presentation will include key takeaways on starting your digital twin journey and ways that Generative AI can provide even greater value throughout the journey, increasing levels of autonomous operation.

Antonios Kontsos, Director Digital Engineering Hub (DEHub) at Rowan University explains that Advanced Manufacturing (AM) provides cyberphysical system approaches to make practically anything, from engineered tissue and cell structures to electronic devices and complex structural materials. To achieve this, AM connects material, process, properties, and applications through digital pathways enabled by advances in data science, computing methods and sensing, creating the basis for interconnecting the physical and digital realms. In this context, Digital Twins (DTs) allow the next level of cyberphysical integration by enabling additional capabilities including automated data processing needed to enable the synchronous and reliable use of physics-based computational and artificial intelligence (AI) models that can be implemented in real-time to assist with fault diagnosis and predict outcomes used to optimize material discovery and synthesis, improve process parameter selection, and assist complex system operation.

As an added benefit of such interconnection, DTs link multiscale information that originates at the single machine level and can expand to encompass enterprise operations, including data-driven lifecycle analysis suitable for circular economy applications. To explain the disruptive changes that this technology landscape is creating, this talk will focus on the process of developing DTs for AM, built to leverage AI, generative AI, and agentic AI coupled with state-of thermomechanical modeling at several hierarchical length scales, multiphysics sensing, and advanced controls to create qualified methods to make anything, while in parallel creating digital pedigrees that interact in interoperable ways with business level decision making.

Sukanya Punthambaker, Co-Founder & CEO Breaking Inc., introduces a biotechnology-driven approach to tackling the accelerating global plastic pollution crisis. With more than 400 million tonnes of plastic produced annually and over 5,000 million tonnes persisting in landfills and oceans, microplastics are increasingly infiltrating ecosystems and even the human body. The company Breaking presents a breakthrough microbial solution built around X-32, a naturally discovered organism capable of degrading multiple types of plastics into harmless byproducts such as CO₂, water, and biomass. Enhanced through synthetic biology, these engineered microbes and enzymes are robust, versatile, and functional across a wide range of environmental conditions. The presentation outlines key applications in wastewater treatment, soil remediation, and ocean cleanup, demonstrating how AI-assisted enzyme development expands the potential for widespread deployment. By leveraging nature-inspired biotechnology and engineered microbial systems, Breaking aims to eliminate microplastic pollution at global scale and restore environmental health.

This presentation introduces a biotechnology-driven approach to tackling the accelerating global plastic pollution crisis. With more than 400 million tonnes of plastic produced annually and over 5,000 million tonnes persisting in landfills and oceans, microplastics are increasingly infiltrating ecosystems and even the human body. The company Breaking presents a breakthrough microbial solution built around X-32, a naturally discovered organism capable of degrading multiple types of plastics into harmless byproducts such as CO₂, water, and biomass. Enhanced through synthetic biology, these engineered microbes and enzymes are robust, versatile, and functional across a wide range of environmental conditions. The presentation outlines key applications in wastewater treatment, soil remediation, and ocean cleanup, demonstrating how AI-assisted enzyme development expands the potential for widespread deployment. By leveraging nature-inspired biotechnology and engineered microbial systems, Breaking aims to eliminate microplastic pollution at global scale and restore environmental health.

Commonwealth Fusion Systems and its partners are in the late stages of building the SPARC tokamak demonstration plant, aiming for net fusion energy production in 2027, and are in the early stages of designing the ARC fusion power plant, aiming for operation in the early 2030s. Fusion energy has the potential to provide large quantities of firm, clean, carbon-free energy in the near term and to continue doing so for the foreseeable future. CFS is building on decades of fusion research from around the world, focusing on a type of fusion plant called a tokamak.

Having demonstrated a new type of high field superconducting magnet in 2021, CFS is now building a high-field tokamak called SPARC, aiming to achieve net fusion energy production in 2027 in a compact, economical facility. In parallel, CFS is also in the early stages of designing a 400 MW electric fusion power plant called ARC, with a site recently announced outside of Richmond, Virginia. This talk will describe the Commonwealth Fusion Systems path to a fusion power plant, where the SPARC and ARC tokamaks sit on this path, and the broader fusion energy ecosystem.

Chris Ritter, Division Director of Scientific Computing & AI and the Director of the Digital Innovation Center of Excellence Idaho National Laboratory, esplains how Idaho National Laboratory (INL) is leveraging AI and ML to streamline the design, development, and licensing processes for nuclear reactors, including automating 3D building information models and facilitating timely regulatory feedback.It will cover INL's advancements in autonomous operations for small modular reactors and microreactors through the development of advanced instrumentation, sensors, and AI algorithms for enhanced safety and efficiency.

The talk also highlights the creation of digital twins for real-time reactor performance monitoring and early anomaly detection. Furthermore, INL's efforts to maximize its specialized test facilities and develop advanced computational models to accelerate research and improve the safety and efficiency of next-generation nuclear power plants will be discussed.

Where science, AI, and nature converge to regenerate ecosystems

Soils are the invisible foundation of our societies. They sustain 95% of our food production, regulate water cycles, and store more carbon than the atmosphere and all vegetation combined. Yet more than half of the world’s soils are degraded, threatening food security, biodiversity, and the resilience of global supply chains. While this crisis is profound, solutions do exist.

Genesis has built the world’s first Soil Health Rating Agency, combining advanced soil science with artificial intelligence and digital platforms. At the heart of this innovation lies the Genesis Soil Reference, the largest global database of soil analyses, which enables a universal and comparable measure of soil health across crops, regions, and practices. This provides the foundation for the Digital Twin of Soil: a powerful tool that models the impact of agricultural practices on fertility, carbon, water, and biodiversity, helping companies and investors anticipate risks, measure resilience, and drive transformation at scale.

During this session, Quentin Sannié, Co-Founder and Chairman of Genesis, highlighted the urgency of restoring soil health, showcase concrete solutions already implemented with leading corporations, and demonstrate how the Genesis platform empowers decision-makers with actionable insights to regenerate ecosystems and secure sustainable supply chains.

Sounds travel very well in the ocean. The different types of sounds are classified from their acoustic sources into three broad categories: biophony, which includes all sounds emitted by living species; geophony, which includes those of the wild environment; and anthropophony, which includes sounds from human activities at sea, considered noise. They disrupt the vital activities of certain species, including cetaceans. Since the 2000s, scientists have characterized the different levels of impacts, ranging from instant discomfort to stranding.

All the individuals of the 90 cetaceans species use their sound emissions during vital activities, such as feeding, reproduction and social interactions. They have created languages ​​based on clicks, whistles, and vocalizations. Analyses have highlighted time sequences, and recent advances in AI make it possible to study larger datasets of acoustic recordings with the ultimate goal of deciphering their languages.

Jan Goetz, Co-CEO and Co-Founder of IQM Quantum Computers, outlines why quantum computing is emerging as the critical next step for addressing humanity’s most computationally demanding challenges. As global issues such as climate change, energy-grid optimization, drug discovery, transportation logistics, and personalized medicine outpace the capabilities of classical computing—and as traditional scaling reaches its physical and environmental limits—quantum computing offers a fundamentally new paradigm.

Goetz highlights IQM’s advances in high-fidelity quantum hardware and tailored algorithmic approaches through concrete case studies: optimizing power plant maintenance with EDF, accelerating photodynamic cancer drug development with Algorithmiq, and achieving chemically accurate quantum-classical simulations of proton transfer in biological systems. These hybrid workflows demonstrate the unique ability of quantum processors to explore complex solution spaces with higher accuracy and efficiency than classical systems.

The presentation positions quantum computing as the next major technological revolution—following mainframes, personal computers, and AI chips—set to unlock transformative value across industries and shape the future of scientific and industrial innovation.

In this presentation, Hon Weng Chong, CEO of Cortical Labs, introduces a groundbreaking paradigm in computing that merges living neurons with silicon to create hybrid biological–digital intelligence systems. Drawing on Moravec’s Paradox, he illustrates how biological neurons—remarkably energy-efficient and capable of rapid adaptive learning—offer capabilities that current AI systems struggle to replicate despite massive computational and power demands.

The talk presents Cortical Labs’ experimental platform, where in vitro human and mouse neurons interface with digital environments, enabling real-time learning and control without backpropagation. Demonstrations, such as task learning and neuron-driven gameplay, highlight the superior fluid intelligence and autonomy of biological substrates compared to deep reinforcement learning models. Chong also outlines the company’s integrated “wetware-as-a-service” architecture, combining custom hardware, software, and neural cultures to build embodied systems that learn and reason dynamically. This work positions biological computing as a transformative new frontier, potentially providing ultra-efficient adaptive intelligence that complements or surpasses traditional silicon-based machine learning.

In this presentation, Barbara Tversky, Professor of Psychology at Stanford University, argues that spatial thinking forms the fundamental basis of human thought. Drawing on extensive evidence from neuroscience, language, gesture, and graphic communication, she shows that the mind evolved to perceive, act, and reason in space long before the emergence of language. Spatial systems in the brain—such as place and grid cells—structure not only our understanding of physical environments but also our representations of events, people, concepts, and abstract ideas.

Tversky demonstrates how gestures, diagrams, maps, and other external representations extend the mind by making thought visible, manipulable, and shareable. These “cognitive tools” enhance comprehension, memory, collaboration, and creativity by transforming fleeting ideas into stable spatial structures. She further explores how ambiguity, perspective-switching, and constructive perception foster new interpretations, supporting innovation across domains from STEM learning to art, design, and problem-solving. Ultimately, the presentation reveals that spatial actions—moving, navigating, arranging—are the foundation on which human reasoning, communication, and creativity are built.

Imagine having a complete map of every type of cell in your body—knowing not just what each cell looks like, but what genes it uses, what functions it performs, and how it communicates with its neighbors. This vision is becoming reality through ambitious cell atlas projects, with the Tabula Sapiens standing as one of the most comprehensive maps of human cells ever created. In this talk, I will take you on a journey from the traditional way we've studied human biology—looking at bulk tissues like examining a smoothie to guess its ingredients—to the revolutionary single-cell approach that lets us examine each "ingredient" individually.

The Tabula Sapiens represents a monumental achievement in biological cartography, cataloging millions of individual cells across 24 human organs from the same donors. This coordination is crucial: for the first time, we can see how a liver cell from one person relates to their heart cell, their brain cell, and their immune cells, creating an integrated view of human cellular diversity. I will explain how we created this atlas using cutting-edge technology that reads the active genetic instructions in individual cells, much like capturing the specific recipe each cell is following at a moment in time. We will explore how this massive dataset reveals that your body contains hundreds of distinct cell types, each with specialized roles that make human life possible.

​​​​​​​An exciting frontier emerging from the Tabula Sapiens is the possibility of creating "virtual cells"—computer models that accurately simulate how real cells behave. Just as weather models help us predict storms by simulating atmospheric conditions, virtual cells could help us predict how our bodies respond to diseases, drugs, and aging. I will demonstrate how researchers are using the Tabula Sapiens data to build these computational cells, starting with the genetic "parts list" from the atlas and adding layers of understanding about how these parts work together. These virtual cells aren't just academic exercises; they're becoming powerful tools for drug discovery, allowing scientists to test thousands of potential treatments on computer models before ever entering a laboratory.

Professor Sunahara received his graduate training with Dr. Philip Seeman in the Department of Pharmacology at the University of Toronto. He later joined the laboratory of eminent biochemical pharmacologist, Dr. Alfred G. Gilman, at the University of Texas Southwestern Medical School as a post-doctoral fellow. His training has provided a strong foundation and appreciation for the applications of pharmacology, biochemistry and structural biology to delineate mechanisms of action.Professor Sunahara started his independent research career in the Department of Pharmacology at the University of Michigan Medical School, where he climbed the academic ladder.In 2015 Professor Sunahara moved his laboratory to the Department of Pharmacology at the University of California in San Diego. His main area of research focuses on the structural and pharmacological bases for hormone-mediated activation of G proteins by G protein-coupled receptors (GPCRs).He has served on several National and International review panels, as well as a peer reviewer and editor for many high-profile scientific journals. He has presented his work at universities, research institutions and scientific conferences all over the world, many as the keynote speaker.He is a Fellow of the American Society of Experimental Therapeutics.

​​​​​​​The Sunahara lab utilizes biochemical, biophysical and pharmacological methodologies to study GPCR-G protein interactions. These approaches were invaluable to resolve the crystal structure of the b2-adrenergic receptor (b2AR)-G protein complex, team effort with long time collaborator Brian Kobilka. The structure was first snapshot of the agonist- and G protein-bound GPCR, providing valuable models for agonist-mediated activation of G proteins. We continue to utilize these data to better understand the basis for receptor-G protein specificity and agonist efficacy. Our mission is to understand the mechanism and structural bases for ligand binding and efficacy to help optimize the design and engineering of more efficacious therapeutics. This is an important perspective in the pursuit of receptor subtype-specific ligands, a major aspect to achieve safer, on-target therapeutics. One example of our recent work surrounds a structure-based effort to develop ligands that specifically target the b2AR above all other adrenergic receptor isoforms. Our goal is to develop safer b2AR-selective ligands for the treatment of asthma and acute rescue therapy for anaphylaxis. We also study non-canonical sites, those outside of the native hormone, or orthosteric, binding sites. We have identified several GPCR ligands that allosterically modulate orthosteric ligand binding and target sites that are often located in regions that display higher sequence variability among receptor subtypes. Again, our intention is to target specific receptor subtypes.

SHIFTBIO is pioneering a new paradigm in advanced therapeutics with our AI-driven Natural Drug Delivery Platform. Conventional drug delivery systems face critical hurdles, including poor target specificity and adverse immunogenicity. To overcome these limitations, we have developed a proprietary platform that transforms cell-derived nanoparticles into a clinically accessible, next-generation drug delivery system, directly addressing previously intractable unmet medical needs.

To accelerate the translation of this innovation to patients, we employ an end-to-end AI solution. This powerful engine integrates big data for target discovery, rational drug carrier design, optimization of scalable manufacturing, and streamlined clinical translation. By leveraging digital simulations, we effectively de-risk and expedite the development of transformative advanced medicines. We will present how our approach creates a tangible bridge between the ambitious vision and the clinical reality of advanced therapeutics, defining both its present and future.

Dr. Gihoon Nam, MD, PhD, is a distinguished physician-scientist and corporate leader. After earning his degrees from Korea University, he completed extensive postdoctoral training in drug development at KIST, Harvard Medical School, and the Dana-Farber Cancer Institute. In 2020, he cofounded and became CEO of SHIFTBIO. As CEO, he leads the company in pioneering next-generation drug delivery platforms, using AI-driven design to accelerate pharmaceutical development under the vision “Make innovation accessible.”

Ariel Stern, Alexander von Humboldt Professor for Digital Health, Economics and Policy at the Hasso Plattner Institute and a Full Professor at the University of Potsdam, examines the rapidly evolving landscape of digital therapeutics (DTx) and software as a medical device (SaMD), highlighting both the transformative potential of software-based medical interventions and the regulatory challenges they pose. She outlines how DTx—software that delivers evidence-based therapeutic outcomes—has expanded across clinical domains, yet continues to face a central question: who will pay for these innovations? Using Germany's pioneering DiGA Fast Track as a case study, Stern illustrates how regulatory frameworks can integrate safety, efficacy, interoperability, and data protection requirements while enabling reimbursement for digitally delivered care. She reviews emerging trends across Europe and the United States, including the FDA's now-concluded Pre-Certification Program, which underscored the need for new legislative tools to regulate dynamic, software-driven medical products.

Stern emphasizes unresolved issues surrounding reimbursement, real-world evidence, adherence, cybersecurity, and long-term software maintenance, noting that current regulatory systems were not designed for rapidly evolving digital products. At the same time, she highlights the significant opportunities offered by DTx, including continuous improvement at scale, lower-cost delivery, and rich data generation to support value-based care. She concludes by calling for deeper understanding of mechanisms of action, improved real-world data collection, and the use of advanced technologies—such as virtual twins and modeling—to unlock the full potential of digital therapeutics for patients and health systems worldwide.

Olivia Caramello is a mathematician working as Associate Professor at the University of Insubria in Como. She is also affiliated to the Université Paris-Saclay. Since March 2022, she has been the President of the Grothendieck Institute, an Italian foundation devoted to interdisciplinary research in the mathematical sciences.

Since the beginning of her Ph.D. studies, her research has focused on investigating the role of Grothendieck toposes as unifying spaces in Mathematics and Logic.

Her main contribution has been the development of the unifying theory of topos-theoretic 'bridges', consisting in methods and techniques for transferring information between distinct mathematical theories by using toposes.

These methodologies are interdisciplinary in character and, even though they find their theoretical and most natural expression in Mathematics, they can also be applied outside it, in subjects such as Physics, Computer Science, Linguistics and Philosophy.

​​​​​​​The Centre for Topos Theory and its Applications of the Grothendieck Institute is notably devoted to developing the theory of toposes as 'bridges' and its applications.

Steve Levine opens the 11th Annual Living Heart Symposium by reflecting on the evolution of the Living Heart Project from a bold idea into a global scientific community and positions the coming decade as a blank canvas for advancing virtual human twins. He highlights the symposium’s role within Science Week, the shift toward precision and continuous digital healthcare, and the importance of collaboration. Nicolas emphasizes that virtual twins are no longer theoretical and are already deployed across industry, research, and healthcare. He frames this moment as a technological and scientific inflection point and invites participants to explore emerging technologies, engage deeply, and help shape the future of human health.

Patrick Johnson presents Dassault Systèmes’ progress in building virtual twins for healthcare, emphasizing investments since the early 2000s in biological and medical modeling. He focuses on cardiovascular disease, oncology, and neurology, all aligned to precision medicine. He describes how MediTwin integrates multimodal data to create virtual representations for prevention, early detection, and treatment planning. He showcases the 3DEXPERIENCE Platform as a scalable infrastructure for advanced modeling, simulation, and AI, and stresses the need for collaboration with clinicians, hospitals, and patient communities to transform healthcare through virtual twin technology.

Michael Hill, Global Head, Science & Technology and Innovation at Science Innovations, makes a patient-centered case that virtual human twins are ready to move healthcare from reactive care to predictive, safer decision-making. He traces the field’s roots from early Physiome work to today’s global initiatives and argues that the science is mature but adoption and trust lag behind. Using Medtronic’s MRI-compatibility simulation program as a proof point, he shows that large-scale in-silico evidence, co-developed with regulators can save years, billions, and patient risk. Hill looks ahead to population-scale and lifelong precision twins but stresses that scalability depends on validated models, clear limits, ethics, and strong security. He closes by challenging stakeholders to build the trust required for virtual twins to transform care.

Kevin Sack, Senior Principal Engineer and Technical Fellow Medtronic, describes Medtronic’s adoption of in-silico, patient-driven development for medical devices. He contrasts traditional reliance on animal tests and preclinical studies with the speed and accuracy achievable through computational modeling. Sack highlights ongoing challenges including prototype scale, organizational inertia, stakeholder diversity, and argues for solutions such as credibility building, scaling modeling resources, integrated communication, and cross-functional decision-making. He illustrates how modeling and simulation are already taking root inside Medtronic to accelerate development and improve outcomes.

Joyce Zhao, Director of Combination Product Device Development at Takeda Pharmaceuticals, presents Takeda’s development and validation of a computational model predicting break-loose glide force in prefilled syringes using FDA’s V&V40 framework. She explains how CFD models hydrodynamic forces and how an FDA model predicts glide force with <1% deviation from bench tests. She shows that all 23 credibility factors meet or exceed rigor goals and emphasizes how the model reduces physical testing, improves efficiency, and ensures patient safety while aligning with ISO 11040 and regulatory expectations.

Dr. David Hoganson discusses the transformative impact of 3D models and computational tools on clinical outcomes at Boston Children's Hospital. He highlights the use of commercial aerospace software to create patient-specific 3D models, improving surgical planning and precision. Specific cases include a patient with congenital heart disease whose operation was guided by a 3D model, and another where computational fluid dynamics predicted the optimal size of a Fontan conduit, leading to a significant improvement in energy loss. Hoganson emphasized the importance of engineering involvement in clinical decision-making, noting that more than 400 models were created in 2024, with plans to create more than 600 in 2025.

Steve Kreuzer explains the advancements and implications of in-silico clinical trials through the ENRICHMENT project, built on Living Heart technology with FDA participation. He highlights V&V40 as the backbone of credible simulation and discusses the transition from preclinical to full clinical applications. He outlines challenges including data access, interdisciplinary communication, regulatory alignment, and ends with a call for deeper collaboration and AI integration to scale future trials.

The panel articulates the importance of in-silico clinical trials in predictive and preventive medicine. Ignacio Lagunas stresses the need for democratization and broad awareness. Kenny Aycock, drawing on FDA experience, explains how simulations reduce uncertainty in device testing. Brandon Freeman highlights risk reduction and cost benefits from early adoption. Steve Kreuzer reinforces the necessity of validated models and standards. Together, they predict widespread adoption and measurable patient outcome improvements within five years.

Speakers

• Steven Kreuzer (Exponent)
• Brandon Freeman (Nexus Healthcare)
• Kenny Aycock (G.RAU Inc.)
• Dr. Ignacio Lugones (Pediatric Cardiac Surgeon, CSO - AVaTAR MedTech, LIU COE)
  
• Moderator : Steven Levine, Dassault Systèmes

Arif Badrou presents UCR’s living lung model integrating experimental and computational methods to analyze lung biomechanics during ventilation. He explains survival improvements when tidal volume is reduced and describes the PV ventilation system for continuous mechanics measurement. He highlights the 3D whole-organ lung model capturing anisotropy and heterogeneity and the team’s move toward patient-specific rat lungs to improve ventilation strategies.

Kristin M. Myers, Professor of Mechanical Engineering Columbia University, New York, describes her work developing digital twins of human pregnancy to prevent preterm birth. She shows why cervical cerclage methods from the 1950s need modernization and presents NIH/NSF-funded research using ultrasound-based modeling to analyze cervical stiffness. Myers identifies the timing of cervical softening as a key risk indicator and discusses studies on twins and uterine scarring. She underscores the enormous human and economic stakes, nearly $1 trillion which digital twins could help address.

Jacob Aptekar, Vice President, Platform AI & Data Science, Medidata Solutions, argues that synthetic data and machine learning are now poised to supercharge virtual twins and clinical trials. He explains how “Simulants”, that are privacy-safe, AI-generated patient twins can unlock Medidata’s massive datasets for cross-institution collaboration. Using CAR-T therapy as a case study, he shows how synthetic clinical twins enable richer virtual cohorts, improved endpoint prediction, optimized enrollment, and faster development timelines. He envisions next-generation simulants powered by transformers and LLMs and outlines how partnerships with regulators can make virtual cohorts trusted clinical evidence.

Steve Levine closes Day 1 with an emotional call to action, describing how virtual twins transform invisible physiology into something clinicians can finally “see.” He revisits the origins of the Living Heart Project, celebrates community achievements, and asserts that virtual twins are shifting from academic promise to real-world impact. He challenges the audience to break the old MedTech paradigm, where half of devices fail, and to embrace virtual twins as a way to invent, validate, and collaborate more safely and effectively.

Nicolas Pecuchet, Director of Corporate Research in Life Sciences and Healthcare at Dassault Systèmes opens Day 2 by affirming that the virtual human twin era has arrived. He highlights evidence from in-silico trials, living organ programs, and real-time predictive modeling. He celebrates the strong interdisciplinary collaboration on display and encourages the community to stay engaged, inspired, and committed to building toward a complete virtual human ecosystem because momentum is accelerating and the opportunity is unique.

Sébastien Fricker, Vision and Perception Modeling Manager EssilorLuxottica, discusses the evolution of ophthalmic lens design, emphasizing the company's innovation and reliance on digital twins. SLR Luxottica, with 200,000 employees and 15,000 patents, has expanded into audiology, wearables, and medical technologies. They use cloud computing to optimize 600 million lenses annually. Fricker details the development of progressive lenses, focusing on wearer-centric design and the introduction of virtual twin technology to simulate real-life scenarios. He also highlights the company's FDA-approved myopia control lenses, which reduce eye elongation by 60-70%, and their efforts to model myopia progression for personalized treatments.

Virtual Twin Experiences are driving the next evolution in medical innovation - enhancing patient care, operational efficiency, and research capabilities. Virtual Twin Experiences enable us to visualize, understand, test, and predict what cannot be seen—from drug effects on disease to surgical outcomes, allowing real-time simulation, predictive analysis, and personalized treatment planning - transforming patient care and improving the lives of patients.

Claire Biot, Vice President, Life Sciences & Healthcare Industry Dassault Systemes, explores how Virtual Twin Experiences are transforming every stage of healthcare delivery — from accelerating drug development and clinical trials to optimizing hospital workflows and patient journeys. By integrating insights across research, clinical, and operational domains, Virtual Twins empower healthcare providers to make data-driven, proactive decisions that enhance efficiency, reduce costs, and most importantly, improve patient outcomes.

Dr. Joseph Rizzo, Harvard Medical School / Massachusetts Eye and Ear, presents new 4D vascular maps of the optic nerve head, revealing far more complex ciliary vessel pathways than classical anatomy shows. These findings explain why optic nerve strokes localize so consistently. Salwa outlines how the team shifts from visualizing anatomy to mechanistic modeling using reduced-order hemodynamic systems capable of handling thousands of microvessels. They note that this strategy could illuminate small-vessel disease in the brain and help analyze lifestyle-linked microvascular risk.

Alon Malka-Markovitz, Research Scientist, Dassault Systèmes Center of Excellence in Digital Engineering, Long Island University, presents advances in living liver twins becoming increasingly realistic and clinically relevant. He highlights the case of a young cancer patient to illustrate unmet needs and describes how patient-specific liver models simulate blood flow, drug interactions, and tissue damage. He explains how the team’s multiscale surrogate models simulate more than 100,000 inlet variations, enabling virtual drug development, improved toxicity prediction, and better surgical planning.

Dr. Jing Bi, SIMULIA Generative Experiences Director, Machine Learning Physics, Dassault Systèmes SIMULIA, demonstrates how combining high-fidelity physics simulation with modern AI enables near-instant, accurate 3D predictions. She outlines Simulia’s parametric-AI workflow and introduces emerging non-parametric, physics-grounded methods that learn directly from real geometries. She showcases examples from automotive and mitral-valve modeling and argues that hybrid physics-AI approaches will accelerate durability prediction, improve surgical planning, and scale patient-specific twins to population-level in-silico trials.

Andrea Falkoff, Vice President of Product Management, Medidata Solutions, presents Medidata’s vision for the clinical trial of the future built on virtual twins, multimodal imaging, and real-world sensor data. She describes automated oncology workflows, external control arms, and radiomics-based forecasts of response and survival. She argues that virtual-twin benefits in clinical care—anticipating outcomes, adverse events, and treatment success—translate directly to trials.

Melissa Ceruolo, VP Engineering & Biomarker Analytics, Medidata Solutions, expands the vision with digital biomarkers and wearable-based functional insights, showing how short monitoring windows provide powerful predictive signals. She emphasizes the need for a robust, interoperable architecture, including multimodal ingestion, semantic knowledge graphs, and clinically trained agents to enable continuous, adaptive trials. 

The MediTwin Showcase demonstrates how virtual twins integrate into real clinical workflows. The session highlights cardiology applications: predicting sudden cardiac death with multimodal patient-specific twins and optimizing complex pediatric heart surgery planning. The team shows how rapid 0D closed-loop models, targeted 3D simulations, and AI-enhanced imaging enable a continuous learning loop from diagnosis to intervention to monitoring.

Speakers

• Pr Damien Bonnet, Head of Cardiology, Necker University Hospital & Imagine Institute, Paris, France
• Aurélien Bustin - Professor of cardiovascular imaging at IHU LIRYC, CHU de Bordeaux, and Bordeaux University
• Cécile Bonnard, Cardiology Twin Technology Senior Manager, Dassault Systèmes
• Hernan Morales, Cardiology Sciences Technology, Dassault Systèmes

The panel shifts focus to training the next generation to use virtual twins and AI as native tools. Alon Malka-Markovic shows how twins make complex phenomena intuitive and expand experiential learning. Christine Myers emphasizes maintaining strong fundamentals and validation discipline. Dr. Joseph Rizzo highlights simulation’s potential to transform microsurgical training and illuminate hidden biological systems. Collectively, they argue for interdisciplinary education rooted in critical thinking, verification, validation, and collaborative problem-solving.

Speakers:

• Arif Badrou (UCR)
• Mohammed Cherkaoui (LIU)
• Kristin Meyers (Columbia)
• ​​​​​​​Joe Rizzo (Harvard)
• Moderator: Joe Baldwin, Dassault Systèmes

Steve Levine, Sr. Director Life Sciences & Healthcare, Dassault Systèmes, closes the symposium with a high-energy call to replace the “valley of death” between research and clinical success using validated virtual human twins that dramatically reduce failure rates before therapies ever reach patients. He argues that pure data-driven AI is powerful but blind, while physics-based twins provide understanding and their fusion is the real breakthrough. He notes strong momentum across FDA, NIH, sensors, and consortium models and stresses that the time is now to connect organ silos into an interoperable whole-body system. He ends with a personal story about his daughter, underscoring why the stakes are real and why the community must act with urgency.