Biomedical engineer Craig Henriquez chases faster compute time.
As a young professor, Dr. Craig Henriquez heard pioneers in electrophysiology estimate that it would take 3,000 years for a computer to simulate 15 percent of a single heart beat in a tissue of one million cardiac cells. That was 1985. Back then, Henriquez, a biomedical engineering professor at the Pratt School of Engineering, was not limited by his imagination, but rather by computing time.
“Computational time was the limiting factor. How many computations you could do per second limited how large the problem you could tackle,” Henriquez said.
By 1998, simulating those one million cells dropped to six or seven hours of computing time. Now it would take minutes.
Henriquez’s research has grown as quickly as computing power has allowed. Today, he relies on vast amounts of computational data to understand how heart arrhythmias occur and how they affect heart function. To do this, he uses mathematical algorithms and computer models to visualize and simulate the heart.
Henriquez’s most recent work on deep brain stimulation is another example of using 3D visualizations and modeling to advance biomedical research.
For much of the 1990s Henriquez relied on a state consortium compute cluster in Research Triangle Park. When it closed down, he worried about the ability to do his research. Duke listened and invested in onsite research computing resources. That investment deepened in 2014 when Duke’s Office of Information Technology enhanced the research computing community with new resources and added support. In 2015, the provost allocated funding for compute cluster vouchers for researchers and set forth a sustained commitment to support research computing for Duke researchers.
Today, Henriquez is tapping into Duke’s Compute Cluster, made up of more than 490 high-performing machines, to power his computational research. He sees patient-specific, real-time 3D heart simulations not far off. This kind of visualization would allow doctors to see and identify heart problems immediately and perform procedures in-hospital using the information. For example, 3D heartbeat iterations could be used to identify regions in the heart that cause arrhythmia, so that they can be precisely treated.
Henriquez has advice for other researchers at Duke. “It’s important to be aware of what’s available at Duke. Have conversations and build relationships with other researchers outside of your field. You don’t have to be a computer science expert, you just need imagination and a willingness to think about how computing can enhance your research.”
Duke researchers in all fields can explore research use-cases, free training and funding opportunities at rc.duke.edu.