Author's School

School of Engineering & Applied Science

Author's Department/Program

Biomedical Engineering


English (en)

Date of Award

Summer 9-1-2014

Degree Type


Degree Name

Doctor of Philosophy (PhD)

Chair and Committee

Sandor J Kovacs


Multiple modalities are clinically used to quantify cardiovascular function. Most clinical indexes derived from these modalities are empirically derived or correlation- based rather than causality based. Hence these indexes don't provide insight into cardiac physiology and the mechanism of dysfunction. Our group has previously developed and validated a mathematical model using a kinematic paradigm of suction- initiated ventricular filling to understand the mechanics of early transmitral flow and the associated physiology/ pathophysiology. The model characterizes the kinematics of early transmitral flow analogous to a damped simple harmonic oscillator with lumped parameters- ventricular stiffness, ventricular viscoelasticity/ relaxation and ventricular load.

The current research develops the theme of causal mechanism based quantification of physiology and uses the kinematic model to study intraventricular fluid mechanics in diastole. In the first project, the role of vortex rings in efficient diastolic filling was investigated. Vortex rings had been previously characterized by a dimensionless index called vortex formation time (VFT). We re- expressed VFT in terms of ventricular kinematic properties- stiffness, viscoelasticity and volumetric preload, using the kinematic model. This VFTkinematic could be calculated using data from a clinical echocardiographic study. The VFTkinematic was a sensitive to physiologic changes as verified by its correlation with a clinically used echo- based index of filling pressure. Additionally, we demonstrated that VFTkinematic, by factoring the ventricular expansion rate, could differentiate between normal filling pattern and pseudonormal filling pattern which is characteristic of moderate DD.

Continuing on our study of intraventricular fluid mechanics, we next studied the development of vortex ring in the ventricle. We discovered that as the vortex ring develops, the leading edge of the circulating flow passes through the main inflow tract. This causes an extra flow wave recorded in transmitral Doppler echocardiography (in addition to early and late filling waves) that had been observed previously. By using cardiac magnetic resonance (CMR) and echocardiography to independently measure intraventricular vortexes we were able to provide a causal explanation for the extra flow wave and its clinical consequences.

We developed another approach to quantify the effect of chamber kinematics on filling via directional flow impedances. In the ventricle, both pressure and flow rate are oscillatory and pressure oscillations cause flow rate changes. Hence a frequency based approach via impedance, to quantify the relationship between pressure and flow rate is intuitive. We developed expressions for longitudinal and transverse flow impedances which could be computed from cardiac catheterization and echocardiographic data. Longitudinal and transverse flow impedances allowed us to quantify the previously observed directionality of filling as a function of harmonics and use it as an index to measure pathophysiologic changes.

While fluid mechanics based indexes provide a method to evaluate LV chamber kinematics in diastole, an alternate approach for DF quantification is LV hemodynamic assessment. Since, LV filling is influenced by pressure changes before and during filling, we investigated the spatial pressure gradient in the LV. We measured the pressure difference between the LV apex and mid-LV using catheterization and we found a larger gradient exists during isovolumic relaxation (2- 3 times) as compared to filling. Additionally, the rate of pressure decay as quantified by different models of relaxation was also significantly different at the two locations.

Additionally, we developed a new method for load independent hemodynamic analysis of the cardiac cycle. Load represents the pressure against which the ventricle has to fill and eject and most LV function indexes are load dependent, which can confound the diagnosis of dysfunction. We computed load independent cardiac cycle hemodynamics by normalizing LV pressure and the rate of change of pressure (dP/dt). Normalization revealed the presence of conserved kinematics during isovolumic relaxation particularly the normalized pressure at peak negative dP/dt while a similar feature was not observed during the contraction. These studies demonstrate the advantage of mechanism based approaches to quantify diastolic physiology.



Permanent URL: