Biomedical engineering

A better way to predict arrhythmia

The cells in the human heart are connected electrically, and the resulting synchronous behavior produces phenomena that have long fascinated Natalia Trayanova. Not all of these phenomena are beneficial, however, and Trayanova, a professor of biomedical engineering at Johns Hopkins, has been studying one that's particularly harmful: arrhythmia. Sudden cardiac death is a leading killer in the industrialized world, and arrhythmias, or abnormal heart rhythms, cause a large proportion of those deaths, according to the American Heart Association journal Circulation. "If you want to make an impact," working in this area of cardiac research "is a place where the impact would be really large," Trayanova says. "That's what I really like."

Together with her team in the Computational Cardiology Lab, she has created a computer simulation that appears to evaluate a heart attack survivor's risk of sudden cardiac death from subsequent arrhythmia more accurately than the current available methods.

Ideally, doctors could use the new computer-based method to determine whether their patients, post–heart attack, should receive a stopwatch-sized implantable cardio defibrillator—ICD, for short—in their upper chest to monitor the heart's electrical activity and deliver a sudden shock should they experience an arrhythmia. Because the scarring left over from a heart attack can disrupt the pathway of the electrical current that propels the beating of the heart, heart attack victims are particularly prone to irregular rhythms.

ICDs have a number of negative side effects, however. First, the devices must be inserted surgically and can result in infection. Second, when they deploy, they deliver an electrical current as strong as a horse kick to the chest, in some cases knocking their owners unconscious. People with ICDs often have to adjust their lifestyles to account for the possibility of a sudden and unexpected shock, and some find themselves suffering harmful psychological effects like acute pain, anxiety, and depression, according to the American Heart Association. Trayanova says some patients endure "fear of the shock, fear of the experience, fear of the pain, fear of the unknown—fear that you might crash if you get it when driving, or drown if you are taking a bath, etc." Additionally, ICDs malfunction and deliver inappropriate shocks often enough for the malfunctions to be termed "common" in a 2011 study published in the Journal of the American College of Cardiology. On an annual basis, only 5 percent of the defibrillators implanted per year ever deliver an appropriate, lifesaving shock, according to a study published in the New England Journal of Medicine in 2005, a finding that suggests the majority of recipients did not need the device in the first place.

One of the main problems, Trayanova says, is that doctors currently use an imperfect indicator to make the call: If the amount of blood the heart pumps out with each beat, known as the ejection fraction, is less than 35 percent, the patient receives the device. If their EF is above 35 percent, they go without. But because the EF is a mechanical measure of how well the heart pumps, it does not accurately reflect the potential for arrhythmia, an electrical dysfunction, Trayanova explains. The other, less common, way to evaluate the risk of arrhythmia is an invasive procedure called an electrophysiological test, which involves inserting a catheter into the chamber of the heart to probe its electrical functioning.

The virtual replica Trayanova developed, by contrast, allows scientists to extensively probe the electrical functioning of the heart under stress—non-invasively—and unmask its problems. The computational lab uses an MRI scan of the patient's heart to create a three-dimensional digital model of the organ that captures the unique shape of its chambers as well as its individualized post–heart attack scarring pattern, the most important determinant of whether an arrhythmia will develop. Then they animate the digital heart on a computer and run a series of test stimuli, in the form of simulated electric currents, through the digital organ, to see if it responds by going into arrhythmia. They've named the procedure VARP—short for virtual-heart arrhythmia risk predictor.

Trayanova ran a retroactive proof-of-concept test on VARP using the MRI scans of 41 Johns Hopkins Hospital patients, all of whom had experienced heart attacks and received surgically implanted defibrillators before 2009. She developed a virtual heart based on each MRI and used VARP to predict each patient's propensity for arrhythmias. Then she compared her findings to the patients' actual medical records. The results were encouraging. Her technique predicted the risk of arrhythmia four to five times better than the traditional method. "The whole benefit of what we are doing is it's personalized," Trayanova says. It's an individualized medical test, not a prediction made from a set of probabilities. "It isn't based on statistics. We are making a decision based on one's personal disease."

Trayanova hopes that once the virtual-heart arrhythmia risk predictor undergoes clinical trials, doctors will begin incorporating the technology into their care for all heart attack sufferers. "I see the mission of my lab as developing a virtual heart for every patient," she says. "We are hoping for a paradigm shift in cardiac patient care down the road."