Public Release: 

University of Pittsburgh researchers develop computer system to predict impending arrhythmia and sudden death

University of Pittsburgh Medical Center

A team of computer software engineers at the University of Pittsburgh, led by Vladimir Shusterman, M.D., Ph.D., UPMC Health System Cardiovascular Institute, has developed and patented a computerized system that can predict a cardiac arrhythmia or sudden death up to eight hours prior to the onset of symptoms. The system is based on the team's research into the general biological mechanisms underlying cardiac arrhythmias and sudden death.

When incorporated into a heart monitor or internal defibrillator, the system can provide ample warning to patients at risk for sudden death that an arrhythmia is imminent and allow time for them to take appropriate action. In the future, the system may be used as an external cardiac monitoring device in hospital intensive care units and has the potential for development as a continuous home-based monitor, according to the researchers. It also may help to prevent unnecessary defibrillator shocks.

"Shocks can be a psychological burden for a person implanted with a defibrillator," said Kelley Anderson, M.D., associate professor of medicine at the University of Pittsburgh School of Medicine, director of Cardiac Electrophysiology Research at UPMC Health System's UPMC Presbyterian Hospital in Pittsburgh and a principal investigator for the clinical part of the study. "If we can predict when an arrhythmia is about to occur, then we can take steps to intervene."

An arrhythmia is an abnormal rhythm of the heart that can be fatal. An internal defibrillator is a battery-powered electronic device implanted near the collarbone of people who are at risk for sudden death. It monitors the heart rhythm continuously and restores normal rhythm by delivering an electrical shock when it detects an arrhythmia. Over 40,000 defibrillators are implanted yearly in the United States.

Some 300,000 people in the United States die suddenly each year from arrhythmia or about one person every one to two minutes. For a large percentage of them, sudden death is their first indication of heart disease.

"Heart rate dynamics are complex and differ from patient to patient. In our research we found that each individual's heart rate has its own traits, like fingerprints," said Dr. Shusterman, an expert in computer technologies and cardiovascular physiology. "The computer system we developed captures these subtle changes in the heart rate pattern."

To develop the program, Dr. Shusterman and colleagues examined EKGs taken from ambulatory heart monitors of over 200 patients with heart disease and cardiac rhythm abnormalities. They examined the dynamics of cardiac cycles, identified the changes that precede initiation of cardiac arrhythmias and developed a mathematical algorithm that detects these subtle changes in electrocardiographic recordings.

"Gradual changes in cardiac cycle dynamics could lead to dangerous arrhythmias and this early prediction could provide a window of opportunity for its prevention," he said.

Researchers found that prediction of atrial and ventricular tachyarrhythmias was accurate 80-90 percent of the time. They could predict the onset of the arrhythmia as long as four to eight hours before the event.

This research is part of a collaborative effort between the University of Pittsburgh and Guidant Corporation, a leading manufacturer of implantable defibrillators.

"Once the computer system is integrated into internal defibrillators, several options are available," said Dr. Anderson. "It could be programmed to preempt the arrhythmia, it can warn the patient to seek medical attention, or it could inject a drug to stop the arrhythmia. The system also has the potential to identify people who may be at risk for sudden death."


Other University of Pittsburgh researchers who participated in this project are: Benhur Aysin, M.S., a Ph.D. student in electrical engineering; G. Bard Ermentrout, Ph.D., Professor of Mathematics; Luis F. Chaparro, Ph.D., associate professor of electrical engineering; and Ilan Grave, Ph.D., assistant professor of electrical engineering.

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.