A Multiscale-Multiphysics Framework for Modeling Organ-scale Liver Regrowth

2024-10-25

arXiv.org

AI Summary

This work addresses the challenge of modeling post-hepatectomy liver regeneration dynamics. We propose the first organ-scale, multiscale, multiphysics model integrating synthetic vascular tree generation, multicomartmental homogenized hemodynamics, poroelastic finite growth, and high-perfusion-driven local volumetric evolution. Crucially, the model couples two key biological mechanisms: perfusion-induced hepatocyte proliferation and “orphaned” vessel–mediated regional hypoperfusion—overcoming a fundamental limitation of conventional regeneration models that neglect perfusion heterogeneity. Leveraging patient-specific geometric registration and effective parameter homogenization, the model accurately reproduces the clinical regeneration timeline—characterized by initial hyperperfusion followed by progressive normalization—at the whole-liver scale, quantitatively matching longitudinal clinical data and precisely identifying ischemia-prone regions. The framework enables preoperative optimization of resection planning and individualized prediction of ischemic risk.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Planning, Routing, and Scheduling: Model-Based Reasoning
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• User Modeling, Personalization and Recommendation: User model development and evaluation
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

We present a framework for modeling liver regrowth on the organ scale that is based on three components: (1) a multiscale perfusion model that combines synthetic vascular tree generation with a multi-compartment homogenized flow model, including a homogenization procedure to obtain effective parameters; (2) a poroelastic finite growth model that acts on all compartments and the synthetic vascular tree structure; (3) an evolution equation for the local volumetric growth factor, driven by the homogenized flow rate into the microcirculation as a measure of local hyperperfusion and well-suited for calibration with available data. We apply our modeling framework to a prototypical benchmark and a full-scale patient-specific liver, for which we assume a common surgical cut. Our simulation results demonstrate that our model represents hyperperfusion as a consequence of partial resection and accounts for its reduction towards a homeostatic perfusion state, exhibiting overall regrowth dynamics that correspond well with clinical observations. In addition, our results show that our model also captures local hypoperfusion in the vicinity of orphan vessels, a key requirement for the prediction of ischemia or the preoperative identification of suitable cut patterns.

Problem

• Modeling liver regrowth
• Multiscale perfusion model
• Poroelastic finite growth model

Innovation

• Multiscale perfusion model
• Poroelastic finite growth model
• Evolution equation growth factor