P7 - Modeling of function-perfusion-deformation interaction on liver lobulus and cellular scale based on a bi-scale continuum FEM model

点击翻译按钮获取中文摘要

The project is part of the research group "Quantification of the liver perfusion-function relationship in complex resection - a systems medicine approach" (QuaLiPerF). Within this research unit, this project aims to numerically simulate the mechanically and biologically coupled perfusion-function processes on the lobular level. The hepatic lobular level is the connection between the next larger (organ vascular system) and the next smaller (cell system) level. This allows the simulation of two- and three-dimensional (2D/3D) liver lobule groups (up to 20). We will study the changes in blood perfusion during fat accumulation and calculate the transient spatial distribution of fat accumulation in the liver lobules. The model will provide information on perfusion changes induced by fat accumulation via portal vein ligation (PVL) and liver resection ((e)PHx). We will extend the model to simulate tissue growth and structural changes during liver regeneration. A time-dependent reorientation approach of the sinusoids will be incorporated. The deformation, flow and transport processes will be modelled using a system of coupled partial differential equations (PDE), while the metabolic processes and fat accumulation will be described using a systems biology approach using a system of ordinary differential equations (PDE-ODE coupling). The high-resolution, hyperelastic and porous lobular model is developed and verified within a thermodynamically consistent continuum mechanical multiphase and multi-scale approach. The model is directly founded on first principles of mechanics and based on the extended theory of porous media (eTPM). In addition, the lobular model is linked to the macro-vessel system at the organ level via coupled boundary values. The variations in blood perfusion and heterogeneity in the liver lobules are transferred to the organ multi-scale model during fat accumulation, after PVL, after (e)PHx and during regeneration. We will increase computational speed and efficiency by approaches to model reduction. Methods for this are the discrete empirical interpolation method (DEIM), proper orthogonal decomposition (POD) or a combination of both. Artificial neural networks (ANN) are to be used as a surrogate model for the high-fidelity multi-phase and multi-scale models.Finally, we will lay the foundation for the long-term vision of the research unit to build a clinically applicable 3D modelling computer tool that enables function-based surgical planning and risk assessment. As a proof-of-concept study, we will visualize clinical and simulation data using a commercial program to create a first demonstration of impaired function and perfusion at organ and lobular level after a predetermined resection.