An Anatomically BasedModelling of Hepatic Perfusion and Biliary Fluid Dynamics in the Human Liver for Clinical Applications

Abstract

This study aims to develop a computational human hepatic perfusion model to predict the pressure and flow rate inside the first few generations of anatomically based hepatic vessels. These vessels were digitised from computed tomography (CT) images of healthy living liver donors. The one-dimensional (1D) model was developed with Navier-Stokes equations and implemented in OpenCMISS, open-source software created by the Auckland Bioengineering Institute. Ultrasonic measurement of blood flow was employed to obtain inlet boundary conditions (BCs). Hepatic perfusion was modelled in the hepatic artery and portal vein with structured tree-based outflow BCs. This self-similar structured tree was used in this study in a novel manner to model the effect of the smaller hepatic arteries and arterioles and the smaller hepatic portal veins and portal venules. As these structured trees terminate at the size of the microvasculature in liver lobules, the structured tree BC has a unique advantage of enabling the proposed organ-level model to be easily connected to tissue-level models of liver lobules. Additionally, blood flow inside the hepatic vein and bile flow inside the intrahepatic bile duct were modelled with RCRWindkesselmodels. The integration of circulation in all hepatic vessels allowed the proposed model to show higher potential for clinical applications. Moreover, the proposed model was applied to predict postoperative hepatic perfusion after left hepatectomy. The resistive index and the characteristic of blood vessel self enlargement to accommodate extra blood flow were employed in a novel manner to obtain proper postoperative BCs for hepatic artery flow and portal vein flow. Additionally, as an informative application, the computational model was coupled with an advection-diffusion equation to model drug delivery phenomena in the hepatic vessels.