Not a member yet? Register for full benefits!

Using a CAVE to Model Blood Flow in Children

CAVEs are a particularly massive form of VR interface installation. They come in several shapes from a cubical room to a star shaped, pointed environment with seven or more walls. Whatever the configuration, all have a simple fact in common: all walls, the ceiling and the floor act as display screens for the same virtual environment, they place one user or a group of users inside the simulation, to interact with it from a single point of view.

There are many advantages to this method. They do-away with avatars for instance, and use the body's own natural movements as a control mechanism. They ensure everyone is seeing the same data at the same time, and they blow up the image to a massive scale. Sometimes as much as seven feet on a side. As a downside however, they are truly massive constructs, requiring six (or more) dedicated projectors, one per surface, and a dedicated minicomputer or small mainframe to handle all the data processing. They are therefore out of the price range of all but the largest universities and dedicated labs.

So, if you are creating a simulation that uses a CAVE, it had better be worthwhile.

One such simulation currently being developed, is certainly that. Two different departments – structural engineers and mechanical engineers - at the University of California, San Diego, are using their newly constructed CAVE installation to collaborate on a novel approach to the human body and prosthetic development.

They are creating blood flow simulations that could lead to improvements in the design of a cardiac pump for children born with heart defects. They hope that the design changes their 3D environment allows them to refine, will improve young patients' outcomes.

Alison Marsden, a professor of mechanical and aerospace engineering, focuses on the development of blood flow simulation tools that can be used to test and optimize new heart surgery designs on the computer before trying them on patients. Yuri Bazilevs, a professor of structural engineering, focuses on computational science and engineering to develop methods for large-scale, high-performance computing applications.

“This work saves a tremendous amount of time, money and risk,” said Dr. Jeff Feinstein, a pediatric cardiologist at Stanford University, who has been working closely with the Jacobs School researchers.

By using the CAVE's capabilities to really get inside the simulation bodily, the researchers have been able to examine the heart, literally from inside, and without relying on the often cumbersome interaction methods of modern avatars. Even better, precisely because there are no avatars to consume computing resources, 100% of the computing power can be focussed on the blood flow simulation.

The work is desperately needed for Americans in particular, because as things stand, the country has only one FDA approved cardiac pump for young children who can’t be outfitted with an adult-sized pump. Whilst this device, the Berlin Heart, is a lifesaver, it is far from perfect. The device carries a 40% risk of developing blood clots, which can lead to strokes or embolisms. This in turn can have devastating consequences on the children using the pump, and as there are no other options, there is no choice but to use it.

Alison Marsden, a professor of mechanical and aerospace engineering, examines some of the simulations her research group developped in the StarCAVE imaging space on the UC San Diego campus


Marsden, Bazilevs and their teams have successfully simulated blood flow within the device. They are now trying to understand how blood clots form inside the pump. The next step is to figure out, through simulations, what design changes are needed to reduce that risk.

The pump has two chambers: one for blood, another for air, separated by a flexible membrane. The air chamber is pressurized, which drives the membrane to pump the blood. But blood flow created by the device is difficult to simulate because of the interaction of blood, membrane and air.

“Blood vessels are complicated,” said Bazilevs, the structural engineer.

“Yuri has been essential in training my team in complex mathematical methods, such as finite element methods, and providing expertise in the development of cutting-edge computational methods,” Marsden said. “I interface with the clinical people to identify high-impact applications, and combined we make a great team.”

He specializes in complex simulations depicting the interaction of several elements. His lab has produced simulations of everything from airflow for wind turbine blades to air flow and water interacting with the hulls of high-speed ships.

"Simulating the heart pump is not simple,” Bazilevs said. “Current commercially available codes are not capable of handling such problems, which necessitate the development and implementation of advanced procedures for the interaction of fluid and structure for this class of applications.”

Marsden’s work focuses on the development of blood flow simulation tools that can be used to test and optimize new heart surgery designs on the computer before trying them on patients. Her group uses sliced based patient imaging data in DICOM format, typically CT, CAT, PET or MRI scans to construct the basic model of the patient's heart and surrounding veins and arteries.

The CAVE's job is then to load the simulation data, using the team's flow algorithms, so the team can try and work out, in near-realtime, what the optimal design would be for that particular patient, to minimise clotting as much as possible, and preferably eliminate it completely. Because the teams have an engineering background, this gives them a curious perspective: most of the tools the simulation uses have been stripped from other simulations, designed to work with CAVE environments to build aircraft.

The basic principles are very similar, and because everything is standardised for this type of interface, it is relatively easy to take the tools used for one simulation and apply them to another. It is just not commonly expected that tools for such different simulations would be useful to one another. It may well be that tools developed in the course of this work, will then go back to the aircraft simulators, for equal benefit there.

It is not all pure research either. So far several of the designs developed in the simulation, have been manufactured by CAD/CAM processes, and then implanted in the same children the medical data came from, by pediatric heart surgeons at Stanford. In addition, she and her students have developed models of heart damage occurring in Kawasaki Disease in collaboration with physicians at the UC San Diego Medical Center and Rady Children's Hospital. These are being put to use in patient treatment.



VR Interfaces: Star CAVE


Computer Simulations Could Lead to Better Cardiac Pump for Children With Heart Defects

Berlin Heart Mechanical Heart Assist

The Berlin Heart (PDF)

Staff Comments


Untitled Document .