Computational and Experimental Investigation of Arterial Hemodynamics

Dynamic arterial blood pressure and blood flow are key determinants of normal or pathological functioning of the cardiovascular system. The measurement of these variables at multiple locations in the body is clinically and physiologically valuable, but difficult to achieve except with invasive methods which carry significant risk to the patient. We have developed and here present a computational model of systemic arterial hemodynamics. The model predicts dynamic pressures and flows throughout the systemic arterial vascular bed. The inputs to the model are pressure or flow measured at a single site, and a description of the architectural and mechanical properties of the blood and blood vessels. We have also measured dynamic pressure and flow noninvasively in healthy women and men. We use these measurements to test and refine the model. The arterial model includes over 24 million blood vessels. The dimensions and branching patterns of 45 large arteries are derived from population averages. Approximately half of these vessels terminate in self-similar branching networks of arteries which extend to capillary-sized vessels. Womersley’s linearization of the Navier-Stokes equations is used to describe the relationship between pressure and flow in each vessel. The inviscid wave velocity in each vessel is estimated based on the combined effects of Young’s modulus, vessel thickness and diameter, and the rheological properties of blood. The blood is modeled as a non-Newtonian fluid whose hematocrit and viscosity vary with vessel size. Wave reflections are computed at all junctions between vessels. The nonlinear pressure drop occurring at the bifurcation of each vessel into daughter vessels is estimated and taken into account when computing the pressures and flows throughout the network. Dynamic pressure is measured noninvasively by applanation tonometry. Dynamic blood velocity is measured with Doppler ultrasonography, and vessel diameter is measured using ultrasound. Custom software uses the electrocardiogram to average data from multiple beats to create ensemble average waveforms for pressure, velocity, and diameter. Data has been collected from the radial and carotid arteries. The experimentally measured pressure from one site is used as input to the model. The model predictions are compared to the other experimental measurements. Blood vessel mechanical properties are estimated by adjusting the model parameters to get good agreement between measured and predicted quantities. This capability can be used to understand effects of pathological changes in vascular properties on local pressure and flow behavior throughout the vasculature.