I am currently leading the Image-Guided Therapy research group at the Medical Imaging Technologies department led by Dorin Comaniciu, Siemens Healthineers, Princeton, NJ (U.S.A). My research area covers medical image analysis (segmentation and registration), computational physiology, in particular cardiac modeling, and artificial intelligence for image analytics and interventional systems.
With the team, we won the R&D 100 award for Siemens eSie Valves, recognized among the top 100 innovative products of 2015. We also won the NJ Thomas Alva Edison Patent Award the same year for our work on personalized simulation of mitral valve interventions. Our research also led to two MICCAI Young Scientists Awards in 2011 and 2013.
I graduated in 2010 from INRIA Sophia Antipolis, Asclepios research team, France.
A complete list of my scientific publications can be found on my Google Scholar profile.
We are hiring! You want to take the challenge to research disruptive technologies for image-guided therapies, going from artificial intelligence, real-time image analytics and computational modeling? Apply and join us!
Inspired by humans, we recently trained an artificial agent to perform the image registration tasks through “Supervised Action Learning”, a technique that combines Deep Learning and Reinforcement Learning under supervision. By just showing examples of aligned images, the agent could figure out how to register CT and DynaCT images together with high performance and robustness. Check out our AAAI 2017 paper, or join us at the conference in San Francisco for our oral presentation!
“An Artificial Agent for Robust Image Registration“, R. Liao et al., AAAI 2017
Check-out our latest paper about Cinematic Rendering, Artificial Agents for Image Understanding, and Deep Learning for Image Fusion and Physiological Computations!
“Shaping the Future through Innovations: From Medical Imaging to Precision Medicine”, Comaniciu et al., Medical Image Analysis, 2016
Designing an inverse modeling approach to estimate the parameters of a computational model of cardiac function can be challenging, in particular when available clinical data is sparse and noisy. Instead, we propose to train an artificial agent exactly like a new user is trained: first learn how the computational model behaves by random and playful experiments. Then let it devise the optimal strategy for parameter identification through reinforcement learning. Applied on several problems, the agent could learn its new task and perform it much more efficiently and accurately than available inverse modeling strategies!
“A Self-Taught Artificial Agent for Multi-Physics Computational Model Personalization“, D. Neumann et al., Medical Image Analysis 2016
The R&D100 awards, often considered as the Oscars of innovation, recognize the “100 most technologically significant products introduced in the past year”. A big big congratulations to the entire eSie Valves team!