(296d) Modeling the Short-Term Effect of Atrial Fibrillation on Hemodynamic Variables | AIChE

(296d) Modeling the Short-Term Effect of Atrial Fibrillation on Hemodynamic Variables


Adeodu, O. - Presenter, Illinois Institute of Technology
Kothare, M., Lehigh University
Gee, M., University of Delaware
Mahmoudi, B., Emory University
Vadigepalli, R., Thomas Jefferson University
Atrial fibrillation (AF) remains the leading cardiac cause of stroke in the United States. An elevated and highly irregular heart rate is the trademark feature of AF and stems from the conduction of abnormal electrical signals from the atria to the ventricles. However, the impact of AF on other hemodynamic variables is not as well established since hypertension, heart failure and other cardiac diseases are often confounding factors. Our goal is to clarify the short-term impact of stand-alone AF on hemodynamic quantities using a lumped parameter approach.

We developed a paroxysmal AF model by making three modifications to a computational model of the human cardiovascular-baroreflex system. Based on the relative constancy of the coefficient of variation of the change in successive heart periods (observed in the MIT-BIH AF dataset), we recast instantaneous heart period as a bounded stochastic variable. In addition, we re-defined the healthy left atrium as a pulsating compartment that boosts ventricular intake during late diastole. This contribution to mitral flow is muted during AF. AF has also been linked to the suppression of parasympathetic drive and a simultaneous enhancement of sympathetic activity. Thus, our third modification involves the development of a valid baroreflex sensitivity metric to quantify the extent of baroreflex impairment.

Our model predictions of the changes in stroke volume and mean arterial pressure during AF episodes match published paroxysmal AF data. Insights obtained from this model will aid the in-silico assessment of potential neuromodulation strategies for AF suppression and also increase the specificity of AF detection algorithms.