Towards the use of geometric parameters and blood-borne species concentration in the prediction of cerebral aneurysm growth and rupture

The rupture of a cerebral aneurysm can have devastating effects, with high rates of morbidity and mortality associated with the event. Approximately 12% of patients who suffer a subarachnoid haemorrhage due to the rupture of an aneurysm die immediately, a further 40% die within one month and one third of those that survive are left with neurological defects. Although the exact mechanisms behind the formation, growth and rupture of aneurysms are unknown, it is believed that a combination of geometric, morphological and haemodynamic factors may play an important role. The primary aim of this research was to investigate the links between geometric features and the mass transport of blood-borne species, and their combined relationship with cerebral aneurysm growth and rupture potential. Improved imaging techniques have allowed for better detection of unruptured cerebral aneurysms. However, the management of unruptured aneurysms, which are estimated to affect 5-8% of the population, is disputed and current intervention methods carry substantial risks for the patient. Currently, the decision to treat patients is based on the likelihood of aneurysm rupture, thus identifying factors that can adequately indicate aneurysm rupture potential is paramount to distinguishing patients at need of surgical or endovascular intervention from those who can undergo observation. However, an issue with many of the parameters previously shown to have a relationship with rupture is that they cannot be directly measured using typical imaging techniques. Consequently, this study investigated the influence of radiologically relevant and measurable parameters on aneurysm rupture in order to determine if geometric features could be used to predict aneurysm rupture potential. A study of 168 patientspecific cerebral aneurysms suggested that aneurysm rupture risk is increased for aneurysms of the posterior circulation, bifurcating in shape, with increasing tilt angle and with increasing aspect ratio. Computational fluid dynamics has become a popular tool for haemodynamic studies of cerebral aneurysms. Current studies focus mainly on factors such as inflow jets, impingement sites, wall shear stress and oscillatory shear index: however, few studies address the influence of the mass transport of blood-borne species on aneurysm rupture. Computational analysis of mass transport of the blood-borne species adenosine triphosphate carried out on a sample set of 12 patient-specific aneurysms of both lateral and bifurcating shape. A species concentration threshold is introduced, which indicates the amount of the aneurysm dome that is exposed to concentrations at 1%, 5%, 10% and 20% less than the inlet concentration, suggesting an underexpression of vital blood-borne species. The computational analysis identified aneurysms that demonstrated areas of reduced concentration as a result of their geometric configuration. Statistical analysis was carried out on the results and rupture probability equations were developed based on the species threshold values, allowing for the prediction of the rupture status of an aneurysm to be made based on the concentration distribution seen within the dome. Results from the geometric and mass transport studies were tested on an independent cohort of ten patient-specific aneurysms, where the rupture status of 8/10 (80%) of the cohort were correctly predicted by the rupture probability equation based on a species threshold of 1%. Furthermore, a combined rupture probability equation, developed based on geometric features and the 1% species concentration threshold led to the rupture status of 7/10 (70%) of the aneurysms being correctly predicted. In previous CFD studies, idealised models were initially used to identify general trends in aneurysm haemodynamics. Idealised models can neglect significant aspects of the geometry such as aneurysmal blebs and irregularities of the aneurysm wall, and so in recent years a shift has been made towards the use of patient-specific models. Patient-specific simulations can be computationally expensive and require large amounts of medical imaging and flow data, and so in this study it was investigated if representative models are a viable alternative. These representative models were developed based on key geometric measurements taken from patient-specific models, and were based on two locations within the CoW: the ophthalmic segment of the internal carotid artery, representing typical lateral shaped aneurysms, and the sylvian bifurcation, representing bifurcation/terminal type aneurysms. Results from mass transport analyses within idealised, patient-specific and representative models demonstrated that while representative models may neglect aspects of the patientspecific geometry such as aneurysmal blebs or undulations, improved information is gained compared to simplified, idealised models in the mass transport of blood-borne species within cerebral aneurysms. Overall, this study indicates that aneurysm rupture risk may be influenced by the amount of the aneurysm dome that is exposed to a species concentration less than 1% of the inlet concentration. The results and conclusions presented throughout this thesis contribute advancements to research in both experimental and computational approaches, increasing our knowledge of the factors that influence cerebral aneurysm rupture potential and may aid in the development of improved aneurysm rupture risk assessment methods, ultimately improving clinical management of this devastating disease.