IBM AND Duke University Team to Study Heart Disease

Duke University researchers will utilize a powerful IBM SP supercomputer to create models of the heart that they hope will lead to uncovering causes and developing treatments for life-threatening heart conditions.

Relying on the same IBM technology used in the U.S. Energy Department's ASCI White supercomputer, Dr. John Pormann and the Electrophysiology research team at Duke University are creating accurate and complex modeling of electrical currents flowing through the heart and nerve tissue.

"Using the IBM SP, the Duke team of researchers can access the horsepower needed for our computationally intense heart modeling," said Dr. John Pormann, Research Associate, Duke University. "The simulations made possible on the supercomputer can give the researchers insight into the problems that generate heart irregularities. These simulations can help provide additional information that is difficult to obtain in the lab."

Irregular heart beats and heart attacks, the leading cause of death in the United States and abroad, are a result of improper electrical impulses flowing through the heart. Complex mathematical computer models, based on lab data,

recreate the heart's reaction to various electrical stimuli. Using the SP supercomputer, researchers can change the model's variables, run simulations and determine the heart's reaction to different electrical stimuli.

For realistic computer modeling of the heart, Duke researchers send huge amounts of data to multiple, ultra-fast processors in the IBM SP supercomputer at the North Carolina Supercomputing Center. With 720 processors, this system is one of the fastest computers in the world -- ranking 16 on the list of Top500 Supercomputers in November 2000.

The IBM SP receives and runs multiple researchers' simulations concurrently on different processors. Researchers can simulate parts of the heart comparing how specific deviations affect the heart function and then incrementally add new complexities to their simulation. Results from various simulations can be compared by running simulations against each other.

As a result of this research, the Duke Computational Electrophysiology Group has developed realistic computer models depicting normal and irregular heart functions.

For more information, visit