Johns Hopkins Engineering Summer Magazine 2013

She’s Got the Beat

Natalia Trayanova, Department of Biomedical Engineering
The heart is much more than the sum of its parts. By capturing that complexity, Natalia Trayanova and her lab aim to deliver clinical innovations that could transform cardiac care.


Written by Jim Duffy
Photography by Marshall Clarke

Once Natalia Trayanova latches onto a goal, she’s not the type to let go of it anytime soon. In fact, she traces her life’s obsession with the natural laws of science to cherished moments spent on her father’s knee back home in Bulgaria. “I remember him reading this book to me about rockets,” she says, traces of her homeland still present in her accent. “It was just so fascinating to me, the laws of physics and how they work when things are going up like that and out of the world. I couldn’t stop asking questions.”

She fantasized as a little girl about becoming an astrophysicist, but it turned out her father, a physiologist, wasn’t done with that business of sharing books that would change her life. During her college years he returned home from an academic conference in the United States with a book titled Bioelectric Phenomena, which explores the way electrical activity functions in nerves and muscles.

“That book was just so amazing to me,” Trayanova says, lifting her head skyward and throwing both arms in the air. “It was the first time I ever really thought about the ways that physics might be applied to biological structures.”

And so Trayanova pulled her sights back from outer space and turned her attention instead to the inner workings of the human body. Today, as the Murray B. Sachs Professor of Biomedical Engineering, she directs the Computational Cardiology Lab (CCL) at the Institute for Computational Medicine, where her team is at the forefront of efforts to understand the intricacies of the workings of the human heart—and to do so in ways that are now showing great potential to help people live longer and healthier lives.

PEOPLE TEND TO THINK about human disease as if it were a linear phenomenon. Sickness has a beginning, and it has an end. It has a cause, and then it has effects. You can plot this way of thinking along a flat chronological line—genetic flaw A leads to biological process B and ends in disease C.

But the human body is not a linear operation, of course, and the mysteries of an organ like the human heart can run as deep in the biological sense as they do in the emotional one.

The landscape where disease unfolds is an astonishingly complex place with genes, proteins, molecules, and cells interacting at every turn as they tackle their tasks. The lines of communication involved don’t just move up from genes toward the organ, but back down as well. And the outside environment, with its pollutants and toxins and other triggers, is always knocking at the door, too.

“There are so many of these interactions going on in the heart, and they are all nonlinear,” Trayanova says. “Even if you know everything that’s happening at the molecular level, you won’t know what that means for the whole heart. The behaviors you end up with at that level of the whole heart often turn out to be much bigger than the sum of the parts.”

Capturing this complexity is what computational medicine is all about. The field traces its roots to the 1950s when British scientists Alan Lloyd Hodgkin and Andrew Huxley developed novel ways to express biological phenomena in mathematical terms. That Nobel Prize–winning work opened the door to creating computer models that mimic biological processes in all their dizzying complexity.

Since arriving in the United States on a National Academy of Sciences fellowship in 1986, Trayanova has focused her modeling work on the heart. Her first post on these shores placed her in the lab of Duke biomedical engineer Robert Plonsey, the man who wrote Bioelectric Phenomena back in 1969. Trayanova went back to Bulgaria briefly after the fellowship but soon returned to the United States and set up her lab first at Duke and then at Tulane.

She came to Johns Hopkins in 2006, one year after the Institute for Computational Medicine was launched as a collaboration between the School of Medicine and the Whiting School of Engineering. Trayanova brought with her a team of 11. Today, 15 researchers are at work in the Computational Cardiology Lab, which receives slightly more than $1 million a year in direct funding and publishes at the rather prolific rate of more than a dozen papers a year.

In building a virtual heart, Trayanova’s team starts by constructing a geometric scaffolding of the organ out of data from magnetic resonance imaging (MRI) scans and computed tomography (CT) scans. Then, they flesh out that scaffolding one piece a time, “populating” the structure with computational representations of the inner workings of the heart, all the way down to cells, molecules, and electrical activity.


Computational Cardiology“We want to be able to see a behavior that’s happening at the level of the whole heart, but also we want to have the ability to zoom in and see the molecular processes and what’s going on in individual cells,” Trayanova says. “I like to describe what we end up with as something like a Google Map—or perhaps a Google Heart.”

Trayanova dubs the heart “the most developed of the virtual organs” and points to structural qualities that make the heart an organ that lends itself to modeling in ways other organs don’t.

“When you think about the brain, the neurons there are connected at the physical points where they meet,” she says. “But in the heart all cells are really connected together, both electrically and mechanically. If something happens in one portion of the heart, very often the rest of the heart will experience something as a result. It’s like what we in our field call syncytium, meaning it’s acting as if it were a single cell.”

Late last year, Raimond Winslow, the Raj and Neera Singh Professor of Biomedical Engineering and director of the Institute for Computational Medicine, was the lead author of a paper in Science Translational Medicine that described this sort of modeling work as a field ready to emerge into the mainstream of medicine. Models such as Trayanova’s heart, stitched together piece by piece over the course of decades, are now putting computational scientists in position to deliver breakthroughs that matter at the patient’s bedside.

“We want to be able to see a behavior that’s happening at the level of the whole heart, but also we want to have the ability to zoom in and see the molecular processes and what’s going on in individual cells. I like to describe what we end up with as something like a Google Map— or perhaps a Google Heart.” Natalia Trayanova

“We are poised at an exciting moment in medicine,” the paper concluded. “Computational medicine will continue to grow as a discipline because it is providing a new quantitative approach to understanding, detecting, and treating disease at the level of the individual.”

Trayanova seems to have a penchant for making a magnetic first impression.

The first time Patrick Boyle met her, he was a graduate student at the University of Calgary, and he was giving his first-ever major presentation at a conference in France.

“I was so nervous, just shaking all over while giving this talk,” he says. “Plus, I was jet-lagged, and I was wearing a suit that was two sizes too small. After I finished, this woman I’d never seen before makes a beeline for me and shouts ‘You must be BOYLE!’ while giving me this big hug. I remember it like it was yesterday.”

Trayanova had heard good things about Boyle from Edward Vigmond, a former postdoctoral fellow of hers who at that time headed the Calgary lab where Boyle was studying. When it came time to apply for postdoctoral work, Boyle had Hopkins at the top of his wish list. He arrived in the fall of 2011.

“The work she’s doing here is very, very groundbreaking stuff,” Boyle says. “And it’s all moving forward because of Natalia’s amazing perseverance. The sheer will she has had over the years to keep pushing forward with this mammoth undertaking is really amazing.”

From the outset, Trayanova’s lab has been focused on modeling strategies that examine the basic mechanisms of heart disease and shed light on what’s going on and why in an ailing heart. But Trayanova has always had her sights set on the step beyond that as well—delivering clinical innovations. That’s an area where her team is now making exciting progress.

“What we want to do is bring computational tools to the bedside,” Trayanova says. “In order to do that, we developed a way of taking MRI and CT scans of patients and using them to construct a Google Heart that’s specific to each individual.”

How might such a patient-specific model be used? Researchers in Trayanova’s lab are currently working with cardiologist Henry Halperin in the School of Medicine to test whether these models can help physicians make better treatment decisions with patients suffering from ventricular tachycardia.

Cardiac ablation, the treatment of choice for this potentially life-threatening arrhythmia, aims to destroy or at least scar the tissue in the heart that’s triggering the problem. But finding the right tissue to target is a challenge. Currently, physicians send patients to the electrophysiology lab for extensive testing with an electric probe. Those tests last between four and 12 hours, but they have a success rate of just 58 percent on the first go-round.

Graduate student Hermenegild Arevalo reports that the patient-specific model predictions of optimal ablation targets beat that success rate by a wide margin in retrospective tests in both pigs and humans. Prospective tests in pigs and humans are now under way.

“The ultimate test for this technique will be trying to do prospective tests in humans,” Arevalo says. “We’re gearing up for that right now.”

A second clinical application aims to better identify which patients need defibrillators because of infarctions that put them at risk of developing arrhythmias. Currently that decision is made by calculating an “ejection fraction”—a measure of how much blood the left ventricle of the heart is pumping. If the number comes in below 35 percent, then the patient receives a defibrillator.


“The problem is that this criterion is very insensitive,” Trayanova says. “Of the patients with implanted defibrillators, only about 5 percent of those defibrillators ever fire because of an arrhythmia. In addition, the criterion misses any patients at risk for sudden cardiac death. We need a better way to stratify these patients.”

Her lab is working with cardiologist Katherine Wu ’89 in the School of Medicine to conduct retrospective tests on patients who have already received defibrillators. The tests are in the early phase, but Trayanova’s researchers have successfully predicted in 13 of 13 cases, to date, whether the defibrillator would fire. Those retrospective tests will continue through about 60 cases before moving to the next phase.

“Much of the work we do is about connecting the dots between the cell, which is heavily studied by one group of people, and the organ, which is heavily studied by a completely different group of people. That’s one of the fascinating things: How do you sort it all out?” Tom O’Hara

Studies with clinical potential are under way in three other areas. Atrial fibrillation is the most common form of arrhythmia and one that puts patients at higher risk for stroke. In patients with fibrotic remodeling in their atria, ablation currently has a very low level of success. Trayanova’s lab is working on predicting what would be the optimal ablation targets in these patients.

In defibrillation, patients receive a painful electric shock that resets the rhythm of the heart to a normal beat. However, in patients with congenital heart disease, a transvenous approach to ICD lead placement is not possible. Trayanova’s lab is working on predicting, in a patient-specific manner, the optimal device placement in these patients. In addition, they are looking to use the patient-specific models to suppress the nerves that carry pain signals while administering the shock.

“The way we envision all of this working in clinical settings, the patients would get scanned and the images would get shipped to a company that specializes in this kind of modeling,” Trayanova says. “It would work just like in a lab test, and they’d come back the next day to say, ‘Here are our predictions for which therapy is best for this individual.’”

While this focus on clinical applications has been generating quite a bit of recent excitement, the Computational Cardiology Lab remains as focused as ever on work that aims to build a stronger base of knowledge about the mechanistic workings of heart disease.

Tom O’Hara came to Trayanova’s lab in the fall of 2011 after earning his doctoral degree from Washington University in St. Louis. His research focus is on finding out what’s going on in individual cells during heart failure.

“There are so many really interesting questions still out there about heart failure,” he says. “We know that older people and fatter people have a higher prevalence of heart failure, but exactly how and why that is true is not actually that obvious. Arrhythmia is something that happens on the organ level, but it’s probably based on things happening on a very small scale.”

O’Hara is working with colleagues at Imperial College London who have developed a novel way to run topographic scans of individual cells. So what happens to the cells in ailing heart?

“The heart cell is typically drawn like it’s a cylinder, but it’s actually more like a sponge,” he says. “There’s all kinds of holes in it that allow the outside environment to communicate with the core of cell.”

These indentations, called transverse tubules (or T-tubules, for short), tend to disappear in heart disease, making the cells less like the sponges they’re supposed to be and more like cylinders.

“There are functional consequences to this at the cell level,” O’Hara says. “My job as a modeler is to take that structural change at the cell level and say, ‘How does that change affect things in the whole heart? How severe—or not severe—are the consequences of this?’ ”

The postdoctoral work of Patrick Boyle focuses on optogenetics, in which low-energy light is used to stimulate and control cell behavior. The field is much further along in neurology, but Boyle is working with colleagues at Stony Brook University to understand its potential in cardiology.

“It’s a new technology, just coming on to the scene,” Boyle says. “And we think it could be used as an experimental tool where we can sort of prompt heart tissue to behave in certain ways so that we can get a better understanding of how arrhythmias come about.”

Down the road, Trayanova sees potential clinical applications in optogenetics as well.

“We are nowhere near there yet with optogenetics,” she says. “But the vision would be this: Instead of using paddles to deliver a shock to the heart, can we accomplish the same thing by shining a light? That would be so much less painful for patients.”

Boyle and O’Hara view the field of computational cardiology as a vital link to the research of colleagues working in more traditional avenues of medical research.

“Much of the work we do is about connecting the dots between the cell, which is heavily studied by one group of people, and the organ, which is heavily studied by a completely different group of people,” O’Hara says. “That’s one of the fascinating things: How do you sort it all out?” Boyle sees modeling as “a way of pulling back the veil.”

“We want to say, ‘OK, now that we know something is happening at the cellular level, what does this mean once we assemble all of the pieces?’” he says. “One of the most important pillars of our work is creating new questions and identifying new experiments for people to try.” Along with the fresh perspective of computational medicine come the fresh eyes of scientists in nontraditional disciplines delving into fields such as cardiology.

And so Trayanova’s lab is as focused on asking questions as it is on answering them. “We span the whole range—from clinical studies to subcellular studies,” Trayanova says. “Our goal is to see how what’s happening at the molecular level manifests itself in the level of the whole heart.”


Tough Love in the Computational Cardiology Lab

Natalia Trayanova, Department of Biomedical Engineering
Kathleen McDowell and Patrick Boyle work in the laboratory under Trayanova’s mentorship.

What’s life like in a lab as prolific as the Computational Cardiology Lab (CCL)? One thing graduate student Kathleen McDowell learned early on is that there’s a trick to scheduling practice sessions for upcoming conference presentations.

“If your talk is going to be 30 minutes long, you’d better schedule the room for two or three hours,” she says.

CCL Director Natalia Trayanova wouldn’t have it any other way. She prides herself on pushing graduate students and postdoctoral researchers to reach their full potential when it comes to presenting findings and defending arguments.

“It’s very important to me that they’re not just good at science but that they can stand on their own and deliver answers when they need to,” she says. “The whole lab participates in practice presentations. It’s grueling. If they can survive that, they’ll be fine.”

Trayanova takes the same approach to grant-writing skills. Five researchers in CCL receive their own grant funding, and pretty much everyone is submitting for grants during their very first year in the lab. The most recent grant that Trayanova herself submitted received the highest score possible from the National Institutes of Health.

“Students write our own papers and our own grants in this lab,” McDowell says. “Her editing process is that she never takes over and rewrites anything—everything is done by comments. So let’s just say there are a lot of iterations involved.”

Postdoctoral researcher Patrick Boyle regards the grant-writing ordeals he’s been through with Trayanova as the experience that might turn out to be most critical to his future success.

“She’s a very demanding person, but what it always comes down to is she wants you to put your best foot forward,” he says. “No matter how much stress she puts you through, it’s about making sure that you’re presenting the true quality of your work and living up to the standards of this lab.”

As the Murray B. Sachs Professor, Trayanova is the only woman to hold an endowed chair in the School of Engineering.

“I’m very big into being a role model in that way,” she says. “I want my girls to be cognizant of the challenges that lie ahead. You should see my girls—let me tell you, these are tough girls, and they can defend their opinions.”

And yet for all the talk of how challenging and arduous the work at CCL can be, graduate students and postdocs rave about the collegial, empowering atmosphere. At some point they all seem to pick the word “family” in searching for a way to describe the place.

“One of the things that sets this place apart is that we’re all such very good friends,” McDowell says. “That goes for Natalia, too. She’s definitely been a little bit of a mother to all of us.”