Human intracranial pulsatility during the cardiac cycle: a computational modelling framework

Abstract

BackgroundToday’s availability of medical imaging and computational resources set the scene for high-fidelity computational modelling of brain biomechanics. The brain and its environment feature a dynamic and complex interplay between the tissue, blood, cerebrospinal fluid (CSF) and interstitial fluid (ISF). Here, we design a computational platform for modelling and simulation of intracranial dynamics, and assess the models’ validity in terms of clinically relevant indicators of brain pulsatility.MethodsWe develop finite element models of fully coupled cardiac-induced pulsatile CSF flow and tissue motion in the human brain environment. The three-dimensional model geometry is derived from magnetic resonance images (MRI), features a high level of detail including the brain tissue, the ventricular system, and the cranial subarachnoid space (SAS). We model the brain parenchyma at the organ-scale as an elastic medium permeated by an extracellular fluid network and describe flow of CSF in the SAS and ventricles as viscous fluid movement. Representing vascular expansion during the cardiac cycle, a pulsatile net blood flow distributed over the brain parenchyma acts as the driver of motion. Additionally, we investigate the effect of model variations on a set of clinically relevant quantities of interest.ResultsOur model predicts a complex interplay between the CSF-filled spaces and poroelastic parenchyma in terms of ICP, CSF flow, and parenchymal displacements. Variations in the ICP are dominated by their temporal amplitude, but with small spatial variations in both the CSF-filled spaces and the parenchyma. Induced by ICP differences, we find substantial ventricular and cranial-spinal CSF flow, some flow in the cranial SAS, and small pulsatile ISF velocities in the brain parenchyma. Moreover, the model predicts a funnel-shaped deformation of parenchymal tissue in dorsal direction at the beginning of the cardiac cycle.ConclusionsOur model accurately depicts the complex interplay of ICP, CSF flow and brain tissue movement and is well-aligned with clinical observations. It offers a qualitative and quantitative platform for detailed investigation of coupled intracranial dynamics and interplay, both under physiological and pathophysiological conditions.

Waterscales: Mathematical and computational foundations for modelling cerebral fluid flow

“…Remark 9. The total pressure formulation of the MPET equations is also advantageous when considering the fluid-structure interaction between the poroelastic brain and the CSF that surrounds it and clinical applications (Figure 4) [6]. The intracranial pressure, a key clinical measure, can be interpreted as the total pressure in the parenchyma and the hydrostatic pressure in the CSF.…”

“…We addressed this goal by designing and introducing mathematical software concepts together with lower-level algorithms for expressing, representing, and solving systems of PDEs coupled across interfaces or subdomains (Figure 8) [10]. These tools enable automated assembly and solution of a wide range of mixed finite element variational formulations, such as, e.g., the finite element spaces and formulations involved in the reduced perivascular flow models (10), interactions across the cell membrane in geometrically resolved models of excitable tissue [16,35] or fluid-structure interfaces [6]. All algorithms are publicly and openly available via the FEniCS Project software [2,10].…”