Degree

Doctor of Philosophy (PhD)

Department

Department of Physics and Astronomy

Document Type

Dissertation

Abstract

Cardiovascular disease is the leading cause of death worldwide. While substantial progress has been made in understanding and managing these diseases, current strategies have not been sufficient to reverse increasing incidence and burden. A potential research solution is the cardiovascular digital twin, a virtual replica of the human circulatory system. However, a digital twin of the entire human vasculature has never been accomplished due to the large computational costs. The goal of this work was to determine the feasibility of a CVDT that includes modeling all vessels in the human body, including physiologically-relevant physical mechanisms. To accomplish this, we used a fractal algorithm to generate all 34 billion blood vessels of the human body, and calculated the time-dependent blood flow using an integrated heart model. We included nitric-oxide-mediated vasodilation, as well as vessel deformation and rupture using peridynamics. To test the computational feasibility, we determined the computational complexity, parallel scalability, and the amount of resources required, including execution time, memory usage, and floating-point operations. We found the CVDT to be computationally feasible. The most computationally expensive task was time-dependent blood flow, which required 500 node-hours, corresponding to a wall-clock time of less than 30 minutes. With further computational optimizations and biophysical improvements, this model has potential to shift the change the paradigm of cardiovascular research and patient care.

Date

6-3-2025

Committee Chair

Newhauser, Wayne

Download

Available for download on Monday, June 05, 2028

Share

COinS