Computer Modeling of a Cerebral Aneurysm

							 Computer Modeling of a Cerebral Aneurysm

Tezduyar’s Group Combines New Algorithms with Rice’s New Cray XD1 to develop Computer
Models of Cerebral Aneurysm
BY JAN E. ODEGARD
CITI Executive Director
Based on advances in computer algorithms for modeling complex fluid mechanics problems, scientists
today are able to simulate, in three dimensions, the flow of blood and the interaction between blood and
the arterial wall. When we combine these new algorithms with access to large- scale computational
resources, we get a glimpse of what might be an emerging tool for enhanced medical diagnostics. Today
flow simulation and modeling can provide far better understanding of the interaction between blood as a
fluid and the flexible blood vessels serving to transport the blood. This enhanced understanding combined
with new computational capabilities may soon add a new tool for the physician, that when combined with
traditional imaging techniques, may soon revolutionize medical diagnostics of aneurysms.
                                     The numerical methods used in the computer modeling described here
                                     were introduced and implemented on parallel computing platforms by
                                     the Team for Advanced Flow Simulation and Modeling (T*AFSM). The
                                     powerful set of numerical methods introduced by the T*AFSM and used
                                     in this computation includes the quasi-direct fluid-structure interaction
                                     method [1, 2] recently developed by the T*AFSM at Rice University.
                                     The computation was carried out on the Cray XD1 supercomputer
                                     recently procured by the Computer and Information Technology
                                     Institute (CITI) at Rice in support of far reaching computational
                                     research.

One of the major computational challenges in cardiovascular fluid mechanics is accurate modeling of the
fluid-structure interactions between the blood flow and arterial walls. The blood flow depends on the
arterial geometry, and the deformation of the arterial wall depends on the blood flow. The mathematical
equations governing the blood flow and arterial deformations need to be solved simultaneously, with
proper kinematic and dynamic conditions coupling the two physical systems.
Using a computed tomography (CT) model of a segment of the middle cerebral artery of a 57 year-old
male, a computer model is able to closely approximate the width and extent of a cerebral aneurysm. The
CT model of the artery approximated in this computation was reported in [3] and the blood flow rate used
in the computation during the systolic cycle is a close approximation to the one reported in [4]. The
computation was carried out on the new Cray XD1 at Rice University. The figure above shows the blood-
flow patterns at an instant during the systolic cycle. The group of three images below shows the arterial
shape at three different instants during the systolic cycle. The computation was carried out by graduate
student Bryan Nanna and research scientist Sunil Sathe, as part the T*AFSM research in cardiovascular
fluid mechanics.




 file:///Y|/Computer%20Modeling%20of%20a%20Cerebral%20Aneurysm%20Final%20-%20filtered.htm (1 of 3)5/8/2006 9:53:24 AM
 Computer Modeling of a Cerebral Aneurysm

                     (a)                                           (b)                                             (c)
                                  Arterial shape at three different instants during the systolic cycle
                   Access to the new Cray XD1 allowed researchers with T*AFSM to carry out simulations
                   significantly faster (3.5 times faster than on an older Intel Xeon based Linux cluster),
                   giving the T*AFSM researchers quick feedback as the they increase the scope and
                   accuracy of their computer modeling techniques. Moving forward, the team will also be
                   able to scale up simulations providing better accuracy and ultimately develop new and
                   better models that can match the conditions more closely, yielding even better insight
                   and understanding. “With access to such large-scale resources locally, we are able to
                   move our computer modeling research forward to new and challenging applications such
                   as cardiovascular fluid-structure interactions,” said Tayfun Tezduyar, James F. Barbour
Professor in Mechanical Engineering and T*AFSM Leader. “The Cray XD1 is about to transform our
research since it permits us to address these new challenges with a level of modeling accuracy previously
not feasible because of a lack of sufficient computational power.”
                             [1]
The Cray XD1, a.k.a. ADA , was acquired with a grant from the National Science Foundation in a
partnership with AMD and Cray. The system, ADA, is host to 336 2.2 Giga Hertz Dual-Core AMD Opteron
                            [2]
processors, a total 1.4 TB of memory, and in excess of 20 TB of disk storage. The system clocks in at
                  [3]
about 3 TeraFlop . The Computer and Information Technology Institute (CITI) at Rice University led this
acquisition from its inception, coordinating the proposal development, system procurement, and
deployment. A team of more than 30 faculty members from Rice spanning Engineering, Natural Sciences,
and Social Sciences participated in the effort to bring this system to Rice to support large-scale computing.
The Computer and Information Technology Institute (CITI) is a research-centric institute dedicated to the
advancement of applied interdisciplinary research in the areas of computation and information technology.

References
      [1] T.E. Tezduyar, S. Sathe, R. Keedy and K. Stein, “Space-Time Techniques for Finite Element
      Computation of Flows with Moving Boundaries and Interfaces,” Proceedings of the III International
      Congress on Numerical Methods in Engineering and Applied Sciences, Monterrey, Mexico, CD-ROM
      (2004).
      [2] T.E. Tezduyar, S. Sathe, R. Keedy and K. Stein, “Space-Time Finite Element Techniques for
      Computation of Fluid-Structure Interactions,” Computer Methods in Applied Mechanics and
      Engineering, 195 (2006) 2002-2027.
      [3] R. Torii, M. Oshima, T. Kobayashi, K. Takagi and T.E. Tezduyar, “Influence of Wall Elasticity in
      Patient-Specific Hemodynamic Simulations,” Computers & Fluids, published online, December 2005.
      [4] R. Torii, M. Oshima, T. Kobayashi, K. Takagi and T.E. Tezduyar, “Computer Modeling of
      Cardiovascular Fluid-Structure Interactions with the Deforming-Spatial-Domain/Stabilized Space-Time
      Formulation,” Computer Methods in Applied Mechanics and Engineering, 195 (2006) 1885-1895


[1]
    The Cray XD1 system at Rice was named after Ada Byron also know as Lady Lovelace and often counted among
the early women pioneers in computing (http://en.wikipedia.org/wiki/Ada_Lovelace)
[2]
    TB is short for terabytes and is a measurement term for data storage capacity equal to approximately 1000
gigabytes (http://en.wikipedia.org/wiki/Terabyte)



 file:///Y|/Computer%20Modeling%20of%20a%20Cerebral%20Aneurysm%20Final%20-%20filtered.htm (2 of 3)5/8/2006 9:53:24 AM
 Computer Modeling of a Cerebral Aneurysm

[3]
    TeraFLOPS: In computing, FLOPS (or flops) is an abbreviation of FLoating point Operations Per Second. This is
used as a measure of a computer's performance, especially in fields of scientific calculations that make heavy use of
floating point calculations. A TeraFLOP is the abbreviation used to denote a trillion FLOPS (http://en.wikipedia.org/
wiki/Teraflop)




 file:///Y|/Computer%20Modeling%20of%20a%20Cerebral%20Aneurysm%20Final%20-%20filtered.htm (3 of 3)5/8/2006 9:53:24 AM