PG Opportunities - Prof T David - UC High Performance Computing - University of Canterbury - New Zealand

Postgraduate Research Opportunities

Research opportunities supervised by Professor Tim David, tim.david@canterbury.ac.nz

Massively Parallel Models for Endothelial/Smooth Muscle Cell coupling in atherosclerotic plaque susceptible vascular geometries

Project Description

It is well known that atherosclerosis occurs at very specific locations throughout the human vasculature, such as arterial bifurcations and bends, all of which are subjected in some part to low wall shear stress (WSS) and significant variations in surface concentrations of blood borne chemical agonists such as ATP. Orientation (axial alignment) and the size of the interstitial space for endothelial cells varies significantly at arterial bends and bifurcations correlating strongly with the position of atherosclerotic plaque formation. Importantly atherosclerotic lesions occur over a much larger length scale than just a single cell and in some cases these lesions grow ‘upstream’ indicating a phenomenon independent of LDL convection.

Little work has been done to investigate the dynamic relationship between relatively large spatial variations in WSS and ATP concentrations on coupled arrays of both endothelial (ECs) and smooth muscle cells (SMCs). Even though nitric oxide (NO) is not considered to be a dominant parameter in the dynamical system the non-linear interaction of heterogeneously activated cells is crucial to understanding both the natural and pathological functions of the endothelium and underlying SMCs in areas of the vasculature predisposed to atherosclerotic plaque formation. Hence

Disruption of inter-cellular coupling between endothelial and smooth muscle cells located in atherosclerotic prone areas of the vasculature (principally arterial bifurcations) is an important factor in atherogenesis.

The neuro/cardiovascular group (led by Prof David), comprising 3 post-graduate students and a post-doctoral position has developed several integrated models where complex blood flow and cellular mechanisms interact. In addition it has been able to utilise these cellular models in replicating the coupled SMC work of a number of international workers.

Physiological Modelling of Neurovascular Coupling using Massively Parallel Computing

Project Description

Local blood flow to the brain is controlled by neuronal activity and special cells called astrocytes. The link between neurons firing and blood flow to the neighbouring brain tissue (termed neurovascular coupling) has become a vibrant area of experimental research over the past five years.

This proposed project will use:

  1. MRI data from our international collaborators in Canada,
  2. validated complex numerical models and
  3. state-of-the-art massively parallel heterogeneous computing techniques to investigate and elucidate the critical cellular chemistry of neurovascular coupling.

There is now evidence that neurological diseases such as Alzheimer’s are fundamentally linked to local blood perfusion of the cerebral tissue. Understanding the relationship between neurovascular coupling and neurological disorders is crucial to providing clinical knowledge and support.

The successful PhD applicant will use distinctive experimental data from our international collaborators in Canada and our unique numerical blood flow models to investigate the role of neurovascular coupling for at least two metabolic “astrocyte” pathways in order to answer the specific research question; which (if any) of these two pathways is dominant in providing local perfusion of brain tissue?

The Brain Research Group (BRG) at the University of Canterbury, led by the Professor David, has developed models of both blood flow in the major cerebrovasculature and complex myogenic/metabolic autoregulation using large tree structures simulating the vascular bed. The tree structures contain millions (up to 10,000,000) segments, with each segment responding to the solution of a set of ordinary differential equations modelling the chemical pathways and subsequent contraction/dilation of arterioles. The auto-regulated cerebral tree model is unique to the BRG and is recognised as both cutting edge and world class.

The computational requirements are substantial and the group in collaboration with IBM researchers at TJ Watson Research Centre and Imperial College have developed a unique heterogeneous supercomputing environment using both the Blue Gene (http://www.hpc.canterbury.ac.nz) and IBM p575 architectures simultaneously to support the solution of the complex chemistry and associated blood flow in both cerebral tissue and the major arteries. The University has just taken receipt of a new 8192 processor Blue Gene/P and a 400 processor IBM Power 755. The student will have access to all supercomputing architectures for this project.

Modelling Arterial Blood Flow in the Brain

Project Description

The Circle of Willis is a ring-like arterial structure located in the base of the brain, and distributes oxygenated blood, to the cerebral arteries. The circle is of particular importance because it allows for blood to be re-routed through the arteries in order to maintain oxygen supply to all of the cerebral territories.

In addition the model contains complex cellular chemistry which simulates the dilation/contraction of the cerebral arterial tree (autoregulation).

An advanced 3D computer model has been developed from Magnetic Resonance Imaging (MRI) data and uses a technique known as Computational Fluid Dynamics (CFD) to model the blood flow throughout the circle of Willis and throughout the cerebro-vasculature. The research involves the use of the UC supercomputer which includes the only Blue Gene in the southern hemisphere. See (www.hpc.canterbury.ac.nz)

The simulation results give the response of the vasculature to different pathological conditions, including the combined effects variations in cerebral-vasculature with atherosclerotic plaque build-up. The long term goal of the project involves the development of a clinical diagnostic tool for automatically recreating a 3D model of an individual’s cerebral vasculature, from an MRI scan and using CFD to predict the likelihood of stroke in the short term, or the risk associated with various surgical procedures such as carotid endarterectomy.

  • UC HPC logo
  • Phone: +64 3 364 2987 ext 45039 (General)
    or +64 3 364 2987 ext 43211 (Technical)
    Fax: +64 3 364 3002

    General Enquiries:
    hpc@canterbury.ac.nz
    Contacts page
  • Postal Address
    UC High Performance Computing
    University of Canterbury
    Private Bag 4800
    Christchurch 8140
    New Zealand
  • Follow us
    FacebookYoutubetwitterLinked In