Randles Wins NIH Director’s Pioneer Award to Predict Heart Disease Risks
By Ken Kingery
Open-ended, $3.5 million award will help tie complex models of blood flow to simple heartbeat measurements through AI
For more than a decade, Amanda Randles has been developing supercomputer models of how blood flows through our arteries at the cellular and particulate level. Dubbed HARVEY for short, the evolving algorithm can simulate how an individual’s specific cardiovascular architecture will likely respond to surgical procedures such as a bypass graft or stent placement.
But depending on the scope of the model, simulating even a handful of heartbeats can require several hours of time on some of the world’s most powerful supercomputers. The sheer volume of data processing needed to investigate 3D cardiovascular dynamics over the course of months or even years makes it intractable for use in informing a patient’s healthcare decisions.
“Single heartbeat metrics have been linked to a person’s risk of heart disease or likelihood of having an adverse event after an intervention, but tracking or predicting how a patient’s blood flow and vascular dynamics respond to a treatment over time is still an open question,” said Randles, the Alfred Winborne and Victoria Stover Mordecai Assistant Professor of Biomedical Sciences at Duke. “We need to find a way around the need for that level of brute force computational power to build a long-term map of a person’s blood flow dynamics across many activities.”
The National Institutes of Health (NIH) agrees and has made Randles one of 2022’s eight recipients of its Director’s Pioneer Award. The Pioneer Award is the largest of the four grants given under the NIH’s High-Risk, High-Reward program and is intended to support innovative research that, due to its inherent risk, may struggle in the traditional NIH peer-review process. Investigators seeking program support are encouraged to think beyond traditional bounds and to pursue trailblazing ideas in any area of research relevant to the NIH’s mission to advance knowledge and enhance health.
“This unique cohort of scientists will transform what is known in the biological and behavioral world. We are privileged to support this innovative science,” said Lawrence A. Tabak, who is performing the duties of the Director of NIH
Randles plans to use the open-ended five-year, $3.5 million award to create physics-based AI models that can estimate a person’s blood flow dynamics over time with the help of wearable devices like smartwatches. By integrating input from wearables with machine learning and physics-based modeling, she hopes to find a way to track changes to a person’s risk of heart disease through indicators such as lowering levels of cardiovascular shear stress.
“By shifting our focus from single heartbeats to creating longitudinal hemodynamics maps, my research program will work to improve both diagnostics and treatment outcomes for patients with cardiovascular disease.”
amanda randles
If successful, the approach could help cut into the $320 billion in annual healthcare costs and lost productivity caused by heart disease and help reduce its mortality rate, as heart disease currently accounts for 25% of all deaths in the United States.
“I’m extremely honored to receive this award and am humbled to be a part of such an esteemed group of recipients,” Randles said. “I’m excited about the new research directions that this award will support and look forward to seeing what we can achieve by combining data from wearables with that from our machine learning and fluid dynamics models. By shifting our focus from single heartbeats to creating longitudinal hemodynamics maps, my research program will work to improve both diagnostics and treatment outcomes for patients with cardiovascular disease.”
Also winning an NIH Pioneer Award is Kafui Dzirasa, the K. Ranga Rama Krishnan Associate Professor of Psychiatry & Behavioral Sciences in the Duke School of Medicine, who holds a secondary appointment in Duke BME. Read more about Duke’s double Pioneer Award win on Duke Today.