Fluid-Structure Interaction Modelling of High-Frequency Wall Vibrations in Intracranial Aneurysms

Abstract

An intracranial aneurysm is an abnormal outpouching of a blood vessel inside the skull, present in over 3% ofpeople. While aneurysm rupture is deadly, many do not rupture, and therefore there is a clinical need for predictors of growth or rupture for unruptured aneurysms discovered incidentally. Since the, 1960s, intracranial aneurysms have been known to produce sounds due to underlying vibrations, however, the mechanism driving the vibrations has been debated, with proposed mechanisms ranging from self-excitation by stable pulsatile flow, to vibration caused by laminar vortex-shedding or turbulent-like flow. This thesis uses recent advances in fluid-structure interaction (FSI) modelling to simulate aneurysm vibrations and reveal the underlying mechanics. Under a slowly increasing, steady inflow, we induced flow phenotypes ranging from stable flow, to harmonic vortex shedding, to turbulent-like flow, in two aneurysm geometries. Both vortex-shedding and turbulent-like flows induced vibrations at specific narrow-band frequencies (100 – 500 Hz) that corresponded with structural mode shapes. We then incorporated pulsatile flow and found that stable pulsatile flow did not induce vibrations, while underlying flow instabilities transmitted through the wall as broad-band, random vibrations (consistent with previously-described bruits), and the sac also exhibited distinct narrow-band vibrations at specific structural modes, which extended into diastole (consistent with previous clinical recordings of musical murmurs). Finally, we showed how aneurysm vibrations could be estimated by un-coupled computational fluid dynamics (CFD) and modal analysis. The modal frequencies observed in aneurysm FSI were predicted accurately by modal analysis, while CFD reasonably predicted the underlying flow instability. Furthermore, the bruit vibration amplitude could be estimated by a linear relationship with the amplitude of flow instability, while the amplitude of the first mode was estimated by a squared relationship with the amplitude of flow instability, combined with the concentration of flow instability at the modal frequency. Our results provide a plausible explanation for distinct intracranial aneurysm sounds, and a pathway towards estimating the likelihood of an individual aneurysm vibrating. Our enriched understanding of the deformation patterns and amplitudes of vibrating aneurysms can inform future mechanobiological experiments to quantify the role of vibrations in aneurysm growth and remodelling.