VIRTUALLY BIOENGINEERING THE LEFT VENTRICLE: A FULLY COUPLED MULTIPHYSICS MODELLING APPROACH
Amr Al Abed, Azam Ahmad Bakir, Nigel H. Lovell, Socrates Dokos
The University of New South Wales, Australia
Computational cardiology is a rapidly evolving field in which in silico models are developed to simulate the function of the heart under healthy as well as diseased conditions. Virtually bioengineered hearts offer the opportunity for in-depth quantitative investigation of mechanisms underlying disease progression and the assessment and development of therapeutic devices and procedures.
The heart’s efficient pumping function is driven by the interaction and synchrony of electrical, solid mechanics and fluid mechanics physics. However, traditional approaches to modelling the heart tend to focus on simulating one or two physics of these components, often limiting the models’ predictive power.
Our group has recently developed a multiphysics electromechanical-fluid cardiac model for simulating the function of the left ventricle (LV) under baseline healthy conditions. The left ventricular geometry was based on a simplified half-ellipsoidal representation. The micro-architecture was formulated such that fiber orientation smoothly changes from -60ᵒ on the epicardium to 60ᵒ on the endocardium. A linear Purkinje fiber network, branching in 3-dimensional space, was incorporated to initiate electrical activation. In addition, the LV model was linked to a Windkessel-type model of the circulatory system. A fully coupled modelling approach was followed to virtually bioengineer the LV function; two-way electro-mechanical as well as twoway fluid-structural interactions were adopted.
The LV function was simulated under a rhythmic heart rate of 60 beats per minute. Despite the simplified geometrical representation of the LV, fiber orientation and the Purkinje network, it was able to reproduce human-realistic epicardial breakthrough times, electrical activation patterns, twisting motion of the ventricle, pressure-volume loop, and ejection fraction. We thus demonstrate the predictive power of our model under baseline conditions and attribute it to our fully coupled implementation of electrophysiology, solid and fluid mechanics. Future studies will utilize this virtual bioengineered LV to simulate pathological conditions and their treatment.