Arteries and Algorithms: Computational physiological flow and arterial disease modelling

Cardiovascular disease, including atherosclerosis, accounts for almost 50% of deaths in the western world. Our understanding of the causes and progression of atherosclerosis is relatively limited and, in many cases, both diagnosis and treatment require invasive patient specific techniques.An enhanced understanding of the physiological factors related to cardiovascular disease would likely to lead to significant advances in treatment as well as increased accuracy in diagnosis and prognosis. Developing this understanding is particularly difficult because of the complexities of the flow in human arterial networks. These are impossible to understand using existing medical data and expertise alone.Recent developments in numerical methods including greater opportunities for wider use of computational simulation and visualisation can provide the necessary link between patient specific imaging data, physics and biology, to provide a platform for this increased understanding. The impact of the use of these techniques could revolutionise medical science and practice in the way imaging modalities such as X-ray, ultrasound and magnetic resonance have done in succession over the past 40 years.The aim of this research initiative is, through close collaboration with vascular biologists, physiologists and surgeons, to develop a simulation environment capable of capturing the multi-scale, hierarchically coupled nature of both physiological and pathological arterial networks. The research programme is focused around three projects, two physiological incorporating the multiscale nature of arterial networks and one numerical: The first project will apply modelling of the three-dimensional fluid dynamics at arterial branches to improve understanding of the causal relationship of blood flow to arterial disease such as atherosclerosis. The second project is directed towards understanding and modelling pulsatile flow wavefoms in patient specific vascular networks using one-dimensional reduced models. The third project focuses on development of advanced mathematical and numerical techniques, such as uncertainty modelling and spectral/hp element methods, to facilitate such modelling.