He built a heart of numbers and equations

The mechanical heart in the fluids lab on the fourth floor of Chapman Hall doesn’t look like a heart. It’s designed to simulate the heart’s mechanics to test medical devices like artificial valves and move devices into clinical testing, but its versatility is limited.

What if a heart made of numbers and equations could supplement its mechanical counterpart by helping researchers test a wider range of parameters with incredible detail and accuracy?

It’s a question Boyce Griffith, a professor and associate chair for research in the mathematics department in the UNC College of Arts and Sciences, has explored since his doctoral training. A uniquely intricate mathematical model of the heart Griffith and his colleagues designed in a recent study won the inaugural Langtangen Prize for research in numerical simulation technology.

“This new model has fully three-dimensional, biomechanically detailed descriptions of all of the heart’s major components, and that allows the model to capture certain physiological features that are hard to capture,” Griffith said.

A strong foundation in mathematics

Growing up near the Oak Ridge National Laboratory in Tennessee, young Griffith had a number of physicists and mathematicians to look up to, including his father, an engineer. He majored in applied math at Rice University, decoding electric signals to learn more about the physiology of neurons.

He enjoyed the work, but he saw a greater potential for the types of systems he could study. For his doctorate at New York University, he turned to the heart.

At first, he was going to simulate the heart’s electrical behavior, but he ended up retooling existing models to run on updated software. This gave him a strong understanding of how fluids and structures interact in the heart.

“The thing I’ve always liked about my work is that, at the end of the day, you have a computer program that predicts what will happen,” he said. “It’s a fascinating concept.”

Attention to detail

Details like the layout of the heart’s muscle fibers can affect the way the organ works, and his group’s models must account for that to make accurate predictions. He strives to simulate not just the motion of the blood and heart muscle but the cause of that motion — the intrinsic mechanical properties of the organ itself.

“The motion of the blood influences the motion of the heart wall, and the motion of the wall influences the motion of the blood and the valves,” Griffith said. “We’re solving the equations to describe all of those things together.”

The result is a simulation that can more accurately forecast how the heart adapts to disease states or medical devices, making predictions that go beyond the data used to train it.

While Griffith and his collaborators have worked toward forecasting individual heart functions for a long time, they only recently developed a model that integrates all the heart’s components into a single simulation. The research, published in the journal PNAS Nexus, was what won the Langtangen Prize.

Griffith has worked closely with colleagues across campus on the project. “I’ve found UNC to be a great environment for the kind of collaborative, interdisciplinary research that was recognized by the prize,” Griffith said. “It’s the thing I always mention first when someone asks me what I like about being at UNC!”

Griffith and colleagues want to continue the collaboration by making their model publicly available, allowing researchers to personalize the code to study different topics. Other groups are using the work to simulate how jellyfish swim and how fluid moves through the esophagus during swallowing.