At Rice University in Houston, Texas, members of the Team for Advanced
Flow Simulation and Modeling (T*AFSM) are working with team members
from other institutions to unlock mysteries of the circulatory system
found within the human brain. The group consists of researchers from
Mechanical Engineering and Materials Science at Rice University, Baylor
College of Medicine in Houston, and Bethel University in Minnesota.
Their computational resources include the computer modeling methods and
programs developed by the T*AFSM and a Cray XD1 supercomputer at Rice.
The interdisciplinary team is using computational fluid mechanics and
fluid-structure interaction to model cerebral arteries, with the goal
of a better understanding—and ultimately, better diagnoses and
prognoses—of cerebral aneurysms.
Leading the effort is Professor Tayfun Tezduyar, James F. Barbour
Professor in Mechanical Engineering at Rice University. Other key
members of the team are Dr. Sunil Sathe from Rice, Professor Brian
Conklin from Baylor, and Professor Keith Stein from Bethel. The
question at hand for Tezduyar and his research colleagues is how
computational science and computer modeling can help surgeons make
treatment decisions about an aneurysm that may be endangering a
patient’s life.
An aneurysm is a weak spot in the arterial wall that begins to bulge
due to pressure, much like a weak spot in an inner tube. It is
estimated that around five percent of people have cerebral aneurysms,
but the simple presence of an aneurysm isn’t necessarily a problem. The
real danger comes from an aneurysm that is in jeopardy of rupturing.
While surgical intervention is possible, operating on a cerebral artery
is highly invasive and comes with very significant risks. So knowing
when—and when not—to operate is a serious question that is often
difficult to answer.
Will It Rupture?
To give surgeons more information about if and when a rupture could
occur, and thus when surgery is required, Tezduyar’s team is simulating
blood flow in the cerebral arteries to identify the blood flow
characteristics and the conditions under which aneurysms evolve and
eventually rupture.
“We want to understand how much an arterial wall with an aneurysm
deforms and how this influences the relevant characteristics of the
blood flow, such as the wall shear stress generated on the arterial
walls,” says Tezduyar.
Understanding exactly how blood flows through and interacts with a
cerebral artery is a major computational challenge. The supple nature
of the artery—and of the balloonlike aneurysm—makes accurate blood flow
simulation very complex. If, for example, an artery were a rigid
structure—like a pipeline with oil flowing through it—blood flow
calculations would be comparatively easy. That’s because the pipeline
can be seen as unchanging in the equations, leaving oil flow
characteristics as the only variable. But that’s not so with blood
flowing through an artery that deforms with changes in blood flow
patterns and pressure.
In the case of an artery, neither the blood flow nor the structure of
the artery can be seen as a constant. In other words, not only does
blood flow change with each beat of the heart, but the supple artery
expands and contracts with it, changing in size and shape continuously.
And when the size and shape of the artery changes, this, in turn,
influences blood flow conditions. For the engineer, this presents what
is known as a fluid-structure interaction, where the two variables
cannot be separated for study because they are constantly influencing
one another.
Engineering Meets Medicine
Before ever becoming involved with biomechanics research, Professor
Tezduyar was interested in designing computational methods and modeling
techniques to solve complex fluid-structure interaction problems. Such
methods and techniques developed by Tezduyar and his graduate students
go back as far as 1992. However, it was only after starting a
collaboration with biomechanics and medical researchers from the
University of Tokyo in 2000 that he became involved in applying his
computational methods to areas outside the traditional realm of
engineering. In fact, the first journal he coauthored on this subject
was in Japanese.
Today, his team is part of a growing trend where researchers of various
disciplines are joining forces to solve problems. As scientists are
able to tackle increasingly sophisticated questions, the techniques and
knowledge required to answer them are often drawn from multiple
disciplines.
Commenting on the trend, Tezduyar points out, “Quite a few people in
the mechanical engineering field are looking at computer modeling of
biomechanics problems. In my estimation, while a lot of the motivation
and direction might be coming from medical schools, most of the
computational methods and know-how are coming from engineering schools.
Of course, those of us who are developing computational methods and
modeling techniques need to work with medical researchers so that we
better understand what needs to be modeled and how to interpret what we
compute.”
A Virtual Aneurysm
To visualize what takes place within the brain, the team began by
creating sections of virtual cerebral arteries with an aneurysm. The
arterial sections were created with preprocessing tools and closely
approximate real arterial sections that researchers imaged with
computer tomography and reported in [1]. Computations are then carried
out in parallel on
the Cray XD1 to simulate numerically how the artery—including the
aneurysm—and blood will interact with one another. This data is finally
fed into EnSight, a software program from Computational Engineering
International (CEI), which allows the blood-artery interactions to be
visualized and analyzed in detail.
One of the areas researchers have been studying through the
visualizations is a type of stress generated on the arterial walls as a
result of blood flow. The existence and extent of this stress, known as
wall shear stress, depend on blood flow characteristics, such as flow
patterns and velocity, and on the geometry of the artery. The
T*AFSM and their Japanese collaborators have found that [1] this is one
area in
particular where it has been critical to take into account the supple
nature of the artery in their calculations.
They found that visual simulations where the artery was assumed to be
rigid gave an incomplete view of wall shear stress’s impact on the
artery. When they considered arterial wall deformation, wall shear
stress results were significantly different. With such realistic
modeling, researchers believe that these visualizations will one day
help them better understand the progression of aneurysms.
“EnSight visualizations have helped us in many ways—and not just for
final presentation media. When we compute something, EnSight also
serves as a diagnostic tool to first make some sense out of our work
and to check our results. For example, we need to make sure that the
laws of physics are accurately represented, and EnSight has good tools
that help us do that. EnSight even gives us vorticity if we want to
have it, so we can look at it and try to understand what is happening
with the true mechanics of the problem. Without EnSight we would have
difficulty making much progress in this simulation,” observes Tezduyar.
The T*AFSM is also creating research and training opportunities for
future generations of US scientists and engineers interested in
computer modeling of biomechanics problems. The T*AFSM members pursuing
research studies in this area include graduate students Tim Cragin,
Bryan Nanna, Jason Pausewang, and Matthew Schwaab and undergraduate
students Arin Lastufka and Will Pryor. As the team continues its
work with cerebral aneurysms, the researchers already have an eye on
future possible endeavors. One area of future work is to continue to
adapt and improve on the numerical methods, for more effective
techniques and more precise simulations. But in addition to this, the
team is planning to next apply its expertise to heart valves.
With the hope of answering some of medical science’s questions and
ultimately improving lives through computer modeling, these researchers
will continue to wend their way through the human circulatory system,
one EnSight visualization at a time.
Blood-flow
patterns at an instant during the systolic cycle.
The computer model is a close approximation of the computer tomography
model [1] of a middle cerebral artery segment of a 57 year-old male
with
cerebral aneurysm.
Blood-flow patterns at an instant during the systolic cycle.
The computer model is a close approximation of the computer tomography
model [1] the middle cerebral artery segment of a 59 year-old female
with a
cerebral aneurysm.
Reference
[1] 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.
Between the Ears
The majority of cerebral aneurysm are congenital, but they can also be
caused through trauma, such as head injuries, or high blood pressure.
Learn more on the basic facts about cerebral aneurysms, such as current diagnostics techniques, risk groups, and more.
Blood-flow
patterns at an instant during the systolic cycle.
The computer model is a close approximation of the computer tomography
model [1] of a middle cerebral artery segment of a 57 year-old male
with
cerebral aneurysm.
