Past Event: Center for Computational Oncology Seminar Series

Patient-specific Modeling of Blood Flow in the Coronary Arteries

Charles Taylor, Ph.D., Founder, HeartFlow, Inc. Adjunct Professor of Computational Engineering and Sciences at the University of Texas, Austin

Abstract

Patient-specific models of blood flow constructed from coronary CT angiography (cCTA) images and using computational fluid dynamics are transforming the diagnosis of heart disease by providing a safer, less expensive and more efficient procedure as compared to the standard of care that often involves nuclear imaging and invasive diagnostic cardiac catheterizations. Such image-based computations require an accurate segmentation of the coronary artery lumen from cCTA images and employ biologic principles relating form (anatomy) to function (physiology). HeartFlow has developed a non-invasive test, FFRCT, based on computing flow and pressure in the coronary arteries [1]. FFRCT has been validated against invasive pressure measurements in more than 1000 patients and demonstrated to improve care in over 100 clinical studies enrolling more than 100,000 patients. At present, FFRCT has been used for routine clinical decision making in more than 250,000 patients in the United States, Europe, and Japan. In the United States, the American College of Cardiology and the American Heart Association guidelines include FFRCT in the recommended diagnostic pathway for heart disease. Medicare and the vast majority of U.S. private insurance companies reimburse physicians for using FFRCT and it is in use in over 850 U.S. hospitals. In England, National Health Services hospitals are mandated to offer this technology to patients. In Japan, the technology is approved for use in over 100 hospitals.  

Patient data is uploaded to the HeartFlow application running on Amazon Web Services and then image analysis methods leveraging deep learning are used to create an initial patient-specific geometric model, which is inspected and corrected by a trained analyst. Next fully-automated mesh generation techniques are used to discretize the model. Computational fluid dynamic analysis is performed on AWS to compute the blood flow solution. Results are returned to the physicians through a web interface or mobile application. New developments including AI-enabled software for quantifying coronary anatomic narrowings, quantifying coronary atherosclerotic plaque, predicting changes in blood flow arising from alternate treatment plans and methods to assess the risk of a heart attack will be discussed.  

References 

[1] Taylor CA, Fonte TA, Min JK., Computational Fluid Dynamics Applied to Cardiac Computed Tomography for Noninvasive Quantification of Fractional Flow Reserve, J Am Coll Cardiol. 2013;61(22):2233-2241. 

[2] Taylor CA, Petersen K, Xiao N, Sinclair M, Bai Y, Lynch S, UpdePac A, Schaap M (2023). Patient-specific Modeling of Blood Flow in the Coronary Arteries. Computer Methods in Applied Mechanics and Engineering. Vol. 417, https://doi.org/10.1016/j.cma.2023.116414. 

Biography

Dr. Taylor is a Founder and Member of the Board of Directors of HeartFlow Inc. He was Chief Technology Officer at HeartFlow from 2010 to 2021 and then Chief Scientific Officer from 2021 to 2023. Prior to HeartFlow, he was an Associate Professor in the Departments of Bioengineering and Surgery at Stanford University with courtesy faculty appointments in the Departments of Mechanical Engineering, Radiology and Pediatrics. He is also currently an Adjunct Professor of Computational Engineering and Sciences at the University of Texas, Austin and a Part-time Professor of Biomedical Engineering at the Technical University of Eindhoven. He is internationally recognized for his pioneering work over the last 30 years in combining computer simulation methods with medical imaging data for patient-specific modeling of blood flow to aid in the diagnosis and treatment of cardiovascular disease. Charles has published over 425 peer-reviewed journal and conference papers and has more than 300 issued patents worldwide. 

He received his B.S. degree in Mechanical Engineering, M.S. degree in Mechanical Engineering and M.S. degree in Mathematics from Rensselaer Polytechnic Institute and a Ph.D. in Mechanical Engineering from Stanford University. Charles was elected into the U.S. National Academy of Engineering in 2024.