Prof. Elie Hachem

 

Biography
Prof. Elie Hachem is a leading expert in computational mechanics and serves as Director of CEMEF (Centre for Material Forming), a world-renowned CNRS research unit at MINES Paris – PSL. CEMEF brings together over 180 researchers, engineers, and doctoral scholars working at the intersection of numerical modeling, physics-based simulation, and industrial innovation.

Prof. Hachem’s research focuses on the numerical development of complex coupled physical phenomena, particularly fluid–structure interaction, conjugate heat transfer, and multiphase flows. He is internationally recognized for pioneering advanced mesh adaptation techniques, and in recent years, for integrating physics-informed machine learning, graph neural networks (GNNs), and reinforcement learning into high-fidelity simulation frameworks.

His work spans domains from smart manufacturing and material forming to cardiovascular therapies, where digital twin technology is transforming personalized treatment planning. He has served as a Visiting Professor at Stanford University since 2012, and is the recipient of major distinctions, including:

- IBM Faculty Award (2015)
- Atos–Joseph Fourier Prize for High-Performance Computing (2019)
- IACM Fellow Award from the International Association for Computational Mechanics (2020)
- ERC Consolidator Grant (2022)
- TwinHeart Industrial Chair (2024)

As a keynote speaker, Prof. Hachem brings deep expertise in translating simulation and AI into real-world impact—at the nexus of engineering performance and human health.

 

Keynote Talk
Digital Twin–Empowered Prediction and Optimization of Vascular Treatments via Space–Time GNNs and High-Fidelity Simulation

Abstract
The integration of high-fidelity simulation with artificial intelligence is unlocking transformative applications in patient-specific medicine. In this lecture, Prof. Hachem presents a cutting-edge computational framework that combines fluid–structure interaction modeling with space–time Graph Neural Networks (GNNs) to build real-time digital twins for cardiovascular treatment planning.

The talk outlines the development of a robust simulation platform that couples blood flow dynamics, vessel wall mechanics, and device interactions with refined temporal resolution. This physics-driven approach is then augmented using space–time GNNs trained to predict transient mechanical and hemodynamic behaviors across diverse patient anatomies.

1Applications are demonstrated in two critical clinical scenarios:

1. Intracranial aneurysm risk assessment, where digital twins provide real-time insights for intervention planning.

2. Heart failure management, featuring optimization of ventricular assist device (VAD) configurations using AI-informed, patient-specific models.

These case studies show how AI-augmented simulation pipelines are revolutionizing both engineering efficiency and clinical decision-making. The methodology aligns with ACTEA’s mission by showcasing how advanced computational tools can reduce cost, improve precision, and enhance the adaptability of engineering systems in both industrial and biomedical contexts.