Numerical Analysis of Oscillatory Flows in the Human Brain by a Lattice-Boltzmann Method

Abstract

The cerebrospinal fluid flow in a brain ventricular system is analyzed by the numerical approach employing a lattice-Boltzmann (LB) method. The cerebrospinal fluid, which surrounds the human brain and spinal cord, fills the cerebral ventricles as well as the cranial and subarachnoid spaces. Diseases in a central nerve system disrupt the flow circulation which influences on a number of vital functions. A computational fluid dynamics technique is used to determine the member geometry impact on the flow motion. The numerical analysis focuses on building a simulation-based basis for testing/optimizing therapeutical methods and understanding the pathophysiology. Magnetic resonance (MR) imaging is exploited to obtain realistic geometries in a brain ventricular system. The computational domain is discretized by a hierarchical Cartesian octree mesh. The numerical procedure based on an LB method overcomes the difficulties raised by typical finite-difference and finite-volume methods on high-performance computing (HPC) systems. An oscillating flow boundary condition is defined to resolve the kinetic behavior of cerebrospinal fluid in a cardiac cycle. The three-dimensional structures captured in the cerebral ventricles show a qualitative agreement with an observation based on an MR velocity mapping. The simulation on a HPC system is able to provide further insights into the transport from brain to spinal cord.

Full document