iSGTW Feature - Unlocking Secrets of the Heart

Feature: Unlocking Secrets of the Heart

A volumetric heart model created from a surface model received from New York University.
Image courtesy of the Computational Visualization Center, University of Texas at Austin

Mathematical models of hearts may not be the most romantic of things, but to patients at risk for vascular disease they hold much more promise than a box of chocolates.

Scientists from the University of Texas at Austin combine images of people's hearts and arteries with mathematical modeling techniques to simulate the interaction between blood flow and the walls of the heart and arteries.

The goal of the project is to help physicians better predict the onset of vascular disease and evaluate treatment plans for affected patients.

"Our work is focused on the development of patient specific models. We use material- and flow-related data from the literature, but by imaging a person, we capture his or her geometry," says Victor Calo, a postdoctoral fellow in the team from the Institute for Computational Engineering and Sciences that is performing the research.

Calo and his colleagues hope to develop tools to retrieve the material and blood flow data from each patient, thus relying less on existing data from previous studies. They also plan to apply these techniques to the modeling and analysis of the human heart.

Their current simulations focus on the flow of blood through the heart and arteries. These studies require solving a complex mathematical problem with many variables-more than can be handled by a single computer processor.

Thus they simulate blood flow using high-performance parallel computers, such as Lonestar, the Texas Advanced Computing Center's Dell Linux Cluster that is available to the national academic research community through the TeraGrid.

To simulate blood flow realistically, the scientists have to include movement of the heart and artery walls in response to the flow. The extra variables introduced by wall movement require vastly more compute cycles to complete, and these cycles are not always available.

"We have proposals coming up in the next month or so to get time on the TeraGrid, and launch our heart-valve modeling project there," says Calo. "We're hungry for more CPU hours to burn!"

Pass those chocolates, would you?

- Anne Heavey, Open Science Grid