Research Papers

Perfusion Characteristics of the Human Hepatic Microcirculation Based on Three-Dimensional Reconstructions and Computational Fluid Dynamic Analysis

[+] Author and Article Information
Charlotte Debbaut

 Biofluid, Tissue and Solid Mechanics for Medical Applications (bioMMeda)Institute Biomedical Technology, Ghent University De Pintelaan 185, Block B, B-9000 Gent, Belgiumcharlotte.debbaut@ugent.be

Jan Vierendeels

Department of Flow, Heat and Combustion Mechanics,  Ghent University, Sint Pietersnieuwstraat 41, B-9000, Gent, Belgium

Christophe Casteleyn

Laboratory for Applied Veterinary Morphology,Department of Veterinary Sciences,Faculty of Pharmaceutical, Biomedical and Veterinary Sciences,  University of Antwerp, Universiteitsplein 1, B-2610 Wilrijk, Belgium

Pieter Cornillie

Department of Morphology,  Faculty of Veterinary Medicine, Ghent University,Salisburylaan 133, B-9820 Merelbeke, Belgium

Denis Van Loo

Centre for X-Ray Tomography, Department of Physics and Astronomy,  Ghent University, Proeftuinstraat 86, B-9000 Gent, Belgium;Department of Soil Management,Ghent University,Coupure links 653, B-9000 Gent, Belgium

Paul Simoens

Department of Morphology, Faculty of Veterinary Medicine,  Ghent University, Salisburylaan 133, B-9820 Merelbeke, Belgium

Luc Van Hoorebeke

Centre for X-Ray Tomography, Department of Physics and Astronomy,  Ghent University, Proeftuinstraat 86, B-9000 Gent, Belgium

Diethard Monbaliu

Department of Abdominal Transplant Surgery,University Hospitals Leuven,  Catholic University Leuven, Herestraat 49, B-3000 Leuven, Belgium

Patrick Segers

Biofluid, Tissue and Solid Mechanics for Medical Applications (bioMMeda),  Institute Biomedical Technology, Ghent University, De Pintelaan 185, Block B, B-9000 Gent, Belgium

J Biomech Eng 134(1), 011003 (Feb 09, 2012) (10 pages) doi:10.1115/1.4005545 History: Received September 23, 2011; Revised December 05, 2011; Posted January 23, 2012; Published February 08, 2012; Online February 09, 2012

The perfusion of the liver microcirculation is often analyzed in terms of idealized functional units (hexagonal liver lobules) based on a porous medium approach. More elaborate research is essential to assess the validity of this approach and to provide a more adequate and quantitative characterization of the liver microcirculation. To this end, we modeled the perfusion of the liver microcirculation using an image-based three-dimensional (3D) reconstruction of human liver sinusoids and computational fluid dynamics techniques. After vascular corrosion casting, a microvascular sample (±0.134 mm3 ) representing three liver lobules, was dissected from a human liver vascular replica and scanned using a high resolution (2.6 μm) micro-CT scanner. Following image processing, a cube (0.15 × 0.15 × 0.15 mm3 ) representing a sample of intertwined and interconnected sinusoids, was isolated from the 3D reconstructed dataset to define the fluid domain. Three models were studied to simulate flow along three orthogonal directions (i.e., parallel to the central vein and in the radial and circumferential directions of the lobule). Inflow and outflow guidances were added to facilitate solution convergence, and good quality volume meshes were obtained using approximately 9 × 106 tetrahedral cells. Subsequently, three computational fluid dynamics models were generated and solved assuming Newtonian liquid properties (viscosity 3.5 mPa s). Post-processing allowed to visualize and quantify the microvascular flow characteristics, to calculate the permeability tensor and corresponding principal permeability axes, as well as the 3D porosity. The computational fluid dynamics simulations provided data on pressure differences, preferential flow pathways and wall shear stresses. Notably, the pressure difference resulting from the flow simulation parallel to the central vein (0–100 Pa) was clearly smaller than the difference from the radial (0–170 Pa) and circumferential (0–180 Pa) flow directions. This resulted in a higher permeability along the central vein direction (kd,33  = 3.64 × 10−14 m2 ) in comparison with the radial (kd,11  = 1.56 × 10−14 m2 ) and circumferential (kd,22  = 1.75 × 10−14 m2 ) permeabilities which were approximately equal. The mean 3D porosity was 14.3. Our data indicate that the human hepatic microcirculation is characterized by a higher permeability along the central vein direction, and an about two times lower permeability along the radial and circumferential directions of a lobule. Since the permeability coefficients depend on the flow direction, (porous medium) liver microcirculation models should take into account sinusoidal anisotropy.

Copyright © 2012 by American Society of Mechanical Engineers
Your Session has timed out. Please sign back in to continue.



Grahic Jump Location
Figure 1

The human liver macrocirculation (left) and microcirculation (right). The liver receives blood from the hepatic artery (HA) and portal vein (PV). The HA and PV branch until they reach the level of the microcirculation at the typical hexagonal lobules (right panel). Every lobule receives blood from its periphery through the hepatic arterioles and portal venules. These are located in the portal triads at each corner of the lobule. Consequently, HA and PV blood is mixed in the sinusoids and drained radially towards the central vein. On their turn, the central veins cluster towards the outflow hepatic veins (HV) and vena cava inferior.

Grahic Jump Location
Figure 2

(a) Scanning electron microscopic image of the sample that was dissected out of the human liver cast. The intertwined and interconnected liver sinusoids are clearly visible in this cast sample representing three liver lobules. (b) 3D reconstruction of the simulation geometry of the case in which a pressure difference is established in the r direction. The origin (0,0,0) is located at the center of the sample. The inflow and outflow guidances are clearly visible, as well as an illustration of the surface mesh density. The meshes existed of approximately 1 × 106 triangular surface elements and 9 × 106 tetrahedral volume elements capturing the sinusoidal geometrical features.

Grahic Jump Location
Figure 3

Dissection of the sample used for the numerical simulations: transectional (left) and longitudinal side (right) views of a hexagonal lobule with indications of the location of the dissected sample. A cube with dimensions (0.15 × 0.15 × 0.15 mm3 ) was dissected of a liver lobule. The dissected cube was oriented in such a way that its z axis was approximately parallel to the direction of the central vein (longitudinal according to the liver lobule). In addition, the r and θ axis were approximately oriented along the radial and circumferential directions, respectively, according to the hexagonal transection of the lobule.

Grahic Jump Location
Figure 4

Static pressure visualization on the boundaries of the three computational fluid dynamics models. The flow direction is from top to bottom (along the r, θ, and z axis, respectively). The top plane was set to a velocity inlet, and the bottom plane to a pressure outlet. The pressure difference for the r and θ simulations are similar within a range of approximately 0–170 and 0–180 Pa, respectively. The z simulation pressure difference is typically smaller, ranging approximately from 0 to 100 Pa.

Grahic Jump Location
Figure 5

Visualization of the streamlines (color coded according to the velocity scale) in combination with the translucent pressure contours. The streamlines clearly indicate the preferential pathways followed by the fluid flow. In addition, the figures clearly illustrate higher velocities at locations where the sinusoids are narrow.

Grahic Jump Location
Figure 6

Wall shear stress distribution color map visualized on the boundaries of the three computational fluid dynamics models. The flow direction is from top to bottom (along the r, θ, and z axis, respectively). Wall shear stresses were typically very low in the major part of all simulation geometries. At the level of narrow channels, wall shear stress increased to higher values.

Grahic Jump Location
Figure 7

Histograms of the wall shear stress distribution on the walls of a cube centrally located in the simulation geometry for each of the three models (flow in the r, θ, and z directions). This corresponds to cube 1 as defined in Table 1. In all cases, wall shear stress was typically in the range of 0 to 1 Pa for the major part of the geometry.




Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In