Design of a Patient Specific Mock Circulatory Loop Using an Elastance Function to Simulate the Frank-Starling Mechanism

Abstract

<h3>Purpose</h3> Mock circulatory loops (MCLs) provide an in vitro test platform to evaluate ventricular assist devices (VADs) for a wide range of clinical conditions. In the human heart, the pressure-volume (PV) relationship in the left ventricle (LV) is an important factor that governs heart performance. Elastance of the LV is time-dependent, and changes in the elastance between systole and diastole are said to drive the cardiac cycle. The purpose of this study is to describe the controller design of an MCL that uses a linear motor-driven pump and a derived LV elastance function template to recreate the heart's PV relationship and simulate the Frank-Starling mechanism. <h3>Methods</h3> The MCL is composed of the left atrium, left ventricle, mitral valve, aortic valve, and a lumped systemic compliance chamber. A closed-loop control system was developed in LabVIEW to control the motion of the piston in the left ventricular simulator using two proportional-integral-derivative (PID) controllers - one controlling systole and the other diastole. The process variable for the PID controllers was the instantaneous left ventricular pressure (LVP), while the output was the voltage feeding the motor controller. Both LVP and left ventricular volume (LVV) were recorded in real time. An analytical PV loop was recreated via the elastance function template and the real time LVV data. The experimental results were then compared to the analytical PV loop. <h3>Results</h3> By comparing the analytical and experimental results, the MCL was able to accurately replicate the PV relationship of the heart. Control using the elastance function allowed for increased cardiac output when preload of the LV increased as characterized by the Frank-Starling mechanism in the heart. <h3>Conclusion</h3> The proposed control algorithm accurately emulated the PV loop produced by the elastance function template. In the future, we will use clinical or preclinical data to derive personalized elastance functions from patients to optimize the MCL's control algorithm.