Numerical study of the cerebrospinal fluid (CSF) dynamics under quasistatic condition during a cardiac cycle

We present a method to perform a numerical simulation of the flow dynamics of cerebrospinal fluid (CSF) based on anatomical magnetic resonance images (MRI). The computational fluid dynamics (CFD) software, written in language C, integrates different numerical schemes to solve the governing equations. The time derivatives were discretized using the Crank-Nicolson scheme. The equation of continuity was modified by introducing an artificial compressibility and discretized by a finite difference scheme. The meshed boundary of the CSF was immersed in a marker-and-cell staggered grid to take into account the fluid-structure interactions. Equations of hydrodynamics were solved with an iterative method under different quasi-static conditions. The anatomical basis of our simulations was generated from individual MRI scans. The surface of the anatomical flow channels of interest was extracted by segmentation and triangulated. In parallel to the acquisition of the anatomical data, CSF flow has been measured by MRI. To characterize a whole cardiac cycle, sixteen equidistant velocity measurements have been performed. In addition, a home made software was implemented to visualize computed data (velocities, pressure).