Scott Hutchinson, Carl Diegert and Louis Romero are using the simulation of electrical current distribution in biological tissues - primarily the heart - to provide the preliminary framework:

MODELING PINPOINTS ANOMALY

For many years, ECGs and vectorcardiograms have been the tools of choice for noninvasive diagnosis of cardiac problems. Through skillful analysis of these skin-surface measurements of cardiac generated electric currents, a physician can deduce the general location of heart conduction irregularities. Sandia's approach would provide information that allows the physician to pinpoint the source of an arrhythmia or an abnormal conduction pathway, such as in reentrant tachycardia or Wolff-Parkinson-White (WPW) syndrome.

POTENTIAL BENEFITS

CORRELATING EXTERNAL MEASUREMENTS TO INTERIOR EVENTS

Picture of a tissue segmentation of my lungs from a set of magnetic resonance images.

Using advanced computational mathematics, Sandia researchers are studying the correlation of the electric current measurements on the patients' body to electrical activity within the heart.

A new MP finite-element code (MPSalsa) developed at Sandia provides efficient numerical solutions to partial differential equation problems. This is combined with Sandia's integrated image processing, solid modeling, and mesh generation capabilities to develop a three-dimensional graphic representation of the voltage and current density distribution on the surface of the patient's myocardium.

IMPROVING TREATMENT TECHNIQUES

Surgical removal of the reentrant circuit is sometimes required for reentrant tachycardia patients who do not respond to antiarrhythmic drugs. Electrocardiographic modeling would greatly simplify the noninvasive location of the reentrant circuit and decrease the time in surgery.

ELECTRICAL DEFIBRILLATION

Picture of finite-element model of current density in and around canine myocardium from defibrillation shock. The three red solids represent regions in the myocardium wall with current density values large enough to cause tissue damage.

Picture of a load balance of the unstructured finite-element mesh of a canine thoracic cavity for distribution on Sandia's massively parallel (MP) supercomputers. The load balance for 32 processors was produced using Sandia's Chaco graph partitioning code. Sandia has collaborated with Dr. Kwong T. Ng at New Mexico State University to apply the same finite-element modeling approach in researching an optimized system that achieves cardiac defibrillation with a minimum amount of electrical voltage.

Anticipated benefits to be gained through use of a lower shock level include:

Optimization is achieved by determining the most efficient electrode configuration. Due to the limited power of conventional computers, previous studies of electrode configurations using numerical techniques have been limited to trial-and-error processes involving a few hand-selected configurations. By using Sandia's new, efficient, parallel finite-element algorithms, a complete and flexible search of feasible electrode configurations can be conducted and a fast, optimal solution can be reached.