Human Arterial Tree Project
Home
Click on the image to view the video. (550KB)
Project Overview
Cardiovascular disease, including atherosclerosis, accounts for almost 50% of deaths in the western world. It is widely accepted that there is a causal relationship between the flow of blood and the formation of arterial disease, such as atherosclerotic plaques, which are observed to occur in regions where both separated and recirculating flows are present, such as vessel branches and bifurcations.
Strong international research efforts into understanding the multi-scale nature of these types of problems, in which interactions among large-scale flow features couple with cellular and sub-cellular biology, are currently underway. Interactions also occur at similar scales in different regions of the vascular system, but this area has received a relatively small amount of research effort due to the high computational demands of modelling these interactions.
High-performance computing is required to model the three-dimensional unsteady fluid mechanics within sites of interest, such as arterial branches and the heart. Experience indicates that capturing the flow dynamics in an artery bifurcation typically requires a mesh of 70,000 to 200,000 spectral elements with polynomial order of 10 to 12. Recent simulations involved 55 main arteries and one million spectral elements.
The Crunch Group developed a version of Nektar utilizing MPICH-G2 and has performed the first cross-site simulations on three NSF TeraGrid sites. More information about Nektar and simulation results are available on the "Software" and "Data" pages, respectively.
Project Goals
Our ultimate goal is to model blood flow interactions in different regions of the cardiovascular system. Currently, our research focus is on two forms of blood flow interactions: first, multi-scale blood flow interactions, such as large-scale flow features coupled with cellular processes, and second, blood flow interactions at the largest scale coupled through pulse wave from heart to the arteries.
Understanding these kinds of interactions is crucial for medical research and practice. Surgical intervention (e.g. bypass grafts) alters wave reflections, which modify wave forms at seemingly remote locations due to coupling. Modification of local wave forms may lead to undesirable wall stress, leading to another disease condition.
We aim to establish a biomechanics gateway on TeraGrid and make the arterial tree a platform and a simulation framework for further developments and systematic studies in hemodynamics, disease modeling, and drug delivery.
