Today, scientists and medical professionals around the world seek a greater understanding of cancer, its causes, and the biological processes driving it. However, due to the complexity of cellular mechanisms, curing cancer remains elusive. While the incredible work of those experts has led to many breakthrough insights and treatments, additional scientific discoveries are needed to identify new ways to attack the devastating disease. Amanda Randles, Ph.D., Alfred Winborne Mordecai and Victoria Stover Mordecai Assistant Professor in the Department of Biomedical Engineering at Duke University is among those researchers taking up the fight. As part of her work, she is preparing to leverage the power of exascale computing to advance cancer research when the Aurora supercomputer is delivered to the U.S. Department of Energy’s Argonne National Laboratory in 2021.
Dr. Randles is renowned in the scientific community for her innovative work developing HARVEY, a computational fluid dynamics model with the extraordinary capability to simulate the blood flow throughout the human body. Since each person’s vascular system includes over 10,000 miles of plumbing – a distance which exceeds that of a round-trip flight from Seattle to Tokyo – that task is arduous indeed.
HARVEY simulation of blood cells and vessels. Credit: Image courtesy of Dr. John Gounley, ORNL
The team spent several years honing and refining the underlying mathematical modeling algorithms enabling HARVEY. They needed to ensure their simulations effectively and accurately simulated blood flow observations in vivo (under real-life conditions) and in vitro (lab-based scenarios).
Ultimately, their endeavor paid off wildly. Today, HARVEY can accurately predict and emulate the path of hundreds of millions of blood cells as they move through the complex maze of the human vascular system. HARVEY can even resolve flow through the smallest capillaries which can be as small as 3-4 microns in diameter. In pipes that narrow, blood cells must travel single-file.
With a deeper understanding of human cardiovascular physiology, Dr. Randles seeks to empower medical professionals who can tap the models for insights that enable more effective treatments for patients. For example, HARVEY could help a doctor identify the ideal location for artery shunt placement or visualize how plaque buildup within arteries impacts a patient’s circulatory health.
HARVEY applied toward cancer research
Not resting on their laurels, Dr. Randles and her teammates at Duke University now seek to use HARVEY in new ways. Because HARVEY can emulate blood flow, the team is refining it for a secondary purpose: Understanding the process of metastasis from cancerous tumors.
Metastasis is a process by which cancerous cells disengage from a primary tumor, hitchhike through the bloodstream, and anchor in another part of the body, potentially seeding a new tumor there. Through this insidious process, kidney cancer, for example, can move from its original tumor and spread to the lungs or brain, making it even more challenging to treat.
Dr. Randles’ work in cancer research resulted from an unexpected encounter. “At a conference, I was approached by another researcher who wondered if we could apply HARVEY to assist in predicting the path of metastatic cancer cells,” recalls Dr. Randles. “We found the idea an important and fascinating challenge with incredible potential. From the beginning, we wanted HARVEY to serve as a tool to assist medical practitioners in the cardiovascular arena. By understanding the biological mechanisms behind metastasized cancer cells too, we hope our work with HARVEY will eventually help doctors and their patients in the fight against cancer.”
Understanding the movement of cancer cells through the bloodstream presents an even more significant challenge than the team’s original work simulating blood flow. Dr. Randles and her team’s augmented model needs to track cancerous cells and demonstrate how they flow through the vascular system, bumping into blood vessel walls all along the way. Also, they must consider how a cancer cell interacts with the multitude of red blood cells as it travels, and how that process impacts where the cancer cell could implant itself in the body.
With the supplemental expertise of Jeffrey Ames, a graduate student at Duke University, HARVEY’s code base is undergoing a critical facelift to support the new endeavor. Argonne researchers Joe Insley, Silvio Rizzi, and Saumil Patel from Argonne National Laboratory, along with Erik Draeger from Lawrence Livermore National Laboratory, are also helping to prepare Randles’ computational studies for the exascale era.
Tapping today’s HPC systems
A mathematically grounded simulation of this complexity requires massive parallel computing capability from some of the world’s most powerful high-performance computing (HPC) systems. So far, Dr. Randles and her team have run HARVEY simulation and modeling with the aid of systems housed at the Oak Ridge National Laboratory (ORNL) and at Argonne.
Even with this stunning level of compute power though, the systems struggle to keep up with HARVEY’s immense workloads involving petabytes of data. “An in vivo simulation of this size requires an HPC cluster with extreme computing power to perform real-time calculations to inform the model. Even on the fastest HPC systems available right now, we use all the available memory to perform the task,” said Dr. Randles. “Tackling our new research into the process of metastasis, and performing the intricate simulations needed, means we need even greater computing power to handle the massive data sets real-time.”
Exascale to the rescue
In 2021, Argonne National Laboratory plans to bring the nation’s first exascale HPC system, dubbed Aurora, online. Aurora’s massive HPC infrastructure will deliver power on an unprecedented compute prowess, on the order of a billion-billion calculations per second thanks to underlying technologies like the future generation of Intel Xeon Scalable processors, Intel Optane DC persistent memory, and technologies based on a new Intel Xe architecture. Added Dr. Randles, “With Aurora’s dexterity bottlenecks of the past will not hamper us.”
Velocity streamlines from HARVEY simulation of flow in a microvasculature geometry from the Kamm Lab at MIT. Credit: Image courtesy of Daniel Puleri, Duke University
To help ensure Aurora is ready to hit the ground running, Dr. Randles, and a small handful of other scientists, are participating in the Argonne Leadership Computing Facility’s Early Science Program to prepare diverse scientific applications for the scale and architecture of the future exascale system. She and her team eagerly anticipate the opportunity to use pre-production time on Aurora to perform more detailed analysis, and at a faster pace, than previously possible.
In addition to the work spearheaded by Dr. Randles, Aurora will enable other advanced research projects like the study of subatomic particles, astrophysics, and global climate analysis.
New insights for cancer treatment
“Initially, our emulation of cancer cell travel will draw upon real-time analysis of a generalized, ‘average’ human vascular system,” noted Randles. However, the team remains focused on bigger, longer-term goals too. Added Dr. Randles, “In the future, we hope to see a day when exascale computing will take things a big leap further. We want doctors to have the ability to examine a unique patient’s cardiovascular system and have better predictive tools to identify where metastasized cancer cells may affix themselves. Doing so will help doctors target their diagnosis and treatment on a per-patient basis – or at least gain insight into the likely spots secondary tumors will form – and watch those areas closely. It’s exceptionally rewarding to come to work each day, recognizing that our research has the potential to save lives.”
Rob Johnson spent much of his professional career consulting for a Fortune 25 technology company. Currently, Rob owns Fine Tuning, LLC, a strategic marketing and communications consulting company based in Portland, Oregon. As a technology, audio, and gadget enthusiast his entire life, Rob also writes for TONEAudio Magazine, reviewing high-end home audio equipment.
This article was produced as part of Intel’s editorial program, with the goal of highlighting cutting-edge science, research and innovation driven by the HPC and AI communities through advanced technology. The publisher of the content has final editing rights and determines what articles are published.