Research work: Humankind is still facing serious diseases that some of them lead to dysfunctions and even to 
shut down of key body organs. Transplantation is yet the main reliable solution.  Unfortunately, the shortage 
of donors and the increasing accumulation of patient recipients in the waiting list represent a very challenging 
problem, which can be tackled by designing artificial organs. Indeed, the latter seriously represents a 
promising alternative. Moreover, downsizing extra and intracorporal medical devices is one of the challenging 
desired features. The goal of the proposed doctoral research project is to contribute to the design of new 
artificial organs that could be fit into microfluidic devices and still mimic efficiently healthy organs. The 
project will use advanced AI/ML and computer simulations to investigate the interplay between the details of 
the microfluidic chip geometry and the performance of organ-on-chip under dynamical flow conditions to 
come up with accurate digital twins of experimentally studied organ-on-chip prototypes at BMBI. 

We will use advanced numerical methods, such as the Lattice-Boltzmann Method for the Computational Fluid 
Dynamics (CFD) and the Immersed Boundary Method for the Fluid-Structure Interaction (FSI). HPC 
simulations will be carried on multiple CPUs to screen a wide range of parameters and to conduct simulations 
with high grid resolutions. Due to the large number of possible parameters to be explored, we consider using 
Machine Learning (ML) techniques as well, and the recent trend Physics-Informed Machine Learning (PIML). 
Such models require a lot of computation power hence the use of parallel computing on GPUs is planned. The 
ML part will be developed in collaboration with Dr. Imad Rida of the C2MUST team.

The developed numerical tools will be used to study blood oxygenation performance of an artificial lung-on-
chip, to analyze how microflows in chips will alter glucose stimulation and insulin secretion of pancreatic islet 
spheroids, and to image process experimental data-set obtained for a platelet-generating bioreactor to 
detect whether megakaryocyte cells are attached or not to bioreactor cylinder pillars. These applications will 
be conducted in collaboration with Dr. Cécile Legallais of the C2B team and Dr. Anne Le Goff of the IFSB 
team.

Key words: Computational Fluid Dynamics, Fluid-Structure Interaction, Machine Learning, Transport 
Phenomena, Artificial organs, Organ-on-chip

Requirements: Strong background in mathematics and physics, with high skills in scientific coding and 
computing (C/C++, Fortran, Python, CUDA, MPI, CAF). MSc in mechanical, chemical or biomedical 
engineering, with an interest in biomedical applications and cutting-edge technologies of computer science 
and engineering.

Starting time: Octobre 2024

Duration: 36 months