Controlling a Heart Simulator Using CompactRIO and LabVIEW

Challenge:

Develop a realistic, reliable and reconfigurable test environment to help the latest cardiac assist devices improve and evolve without the need for animal testing.

Solution:

Create a standalone hardware-in-the-loop (HIL) test environment using NI CompactRIO. This test environment combines an artificial mechanical heart with a circulatory blood flow model to create a living solution with a realistic hemodynamic environment.

Heart disease causes nearly half of all deaths in developed countries. Heart transplants remain the most effective treatment for heart disease, but the number of donated organs far exceeds the demand. To address this imbalance, research is underway to use a novel mechanical artificial heart assist device being developed at the University of Leeds called the intelligent ventricular assist device (iVAD). The device acts as an artificial muscle wrapped around the heart to assist a failing heart by applying pressure around the outer surface of the heart's ventricles in sync with the heart's natural rhythm. This cyclical "squeezing" action increases myocardial power and improves blood output in a diseased heart.

We needed to realistically apply the iVAD to a simulated heart in order to measure the effects of pressure on it, so a realistic in vitro test environment was imperative for development. In the past, other heart assist device testing systems have typically used large mechanical simulated circulatory systems or used isolated hearts that were supported by the blood circulation of another animal. Neither approach was practical for us, so we created a unique HIL (hardware-in-the-loop) heart simulator that combines a real-time software blood flow model with a physical 3D artificial heart. We further enhanced the test environment using the NI LabVIEW graphical programming environment and CompactRIO so that the heart simulator can function like a stand-alone system and operate reliably for longer durations.

Heart Simulator Principles

We needed a heart simulator that could be reconfigured to replicate the real blood environment of different patient types, disease types, and animal models. This adaptation would reduce reliance on animal testing because the heart simulator could extend testing with the iVAD prototype and provide information about the physiological effects of the iVAD. For assist

devices such as the iVAD, the interaction between the assist device and the heart surface is critical. This interaction is likely to depend on difficult-to-simulate human characteristics such as clearances and nonlinear friction; therefore, it was critical for the heart simulator to have a physical object that the iVAD could interact with so that we could monitor raw data during compression.

Heart Simulator Design

In designing the heart simulator, we used the principles of HIL simulation. This is a common testing technique in industry. HIL simulates some components of a system in software and connects them to specific real hardware in the same system that needs to be tested through I/O. To meet the requirements of the heart simulator, we used a mechanical heart as the hardware part of the HIL simulation and placed it in a simulated blood flow circulation model. The continuous interaction loop between the two was evaluated to understand how the iVAD can assist and affect the heart and blood flow when implanted in the human body.

The shape of the artificial heart is determined by two deformable semicircular structures, which are composed of bent spring steel bars. The steel bars are fixed at both ends and their boundary shapes are adjustable. We also developed a custom NI vision program to determine the necessary boundary shapes to match the contour of each steel bar with the reference heart model. We used two linear actuators to achieve cyclic control of the bent steel bars to realistically represent the dynamic movement of the left and right ventricles of the heart. We controlled the motion of the actuators in the blood flow model to simulate the motion of the simulated heart, so any volume changes of the simulated heart would directly affect the artificial heart. In addition to being able to match the shape of the heart, this design also allowed us to change the local stiffness of the artificial heart periphery by changing the mechanical properties of the steel strips (such as thickness) individually. Finally, we wrapped a thin layer of elastic band around the steel strips to implement the iVAD.

Heart Simulator Implementation

As mentioned above, we used a loop with feedback to evaluate the cardiovascular assistance of the iVAD. Four similar pressure sensors were placed at equal intervals around the artificial heart to provide data during the iVAD assistance process (compression process). Within the model, this data was converted into assistance pressure for each ventricle, and the subsequent impact on blood flow was calculated in real time, and finally output to the hardware and the movement of the artificial heart was changed accordingly. The

blood flow model works similarly to a closed-loop lumped parameter model of an electrical network. Because each area of ​​the heart is simulated separately, we can achieve local control of the heart and adjust for specific heart conditions or heart diseases. To meet our primary goal, the blood flow model can be automatically adjusted to represent physiological data using a nonlinear least squares parameter estimation method (implemented as a state in the LabVIEW code). This means that the heart simulator can accurately reflect the hemodynamic characteristics of most pathologies and in vivo models, helping to improve our understanding of the potential effects of the device.

We use CompactRIO to control the artificial heart, run the simulation, and send data via TCP to the Windows host for display and storage. The real-time controller can execute two loops running in parallel: a high-priority control loop to control the blood flow model, and a low-priority communication loop that can send and receive queued TCP data to the Windows host. The high-priority blood flow model loop runs at 500 Hz and converts the two ventricular volumes into calibrated positioning voltages. The positioning voltages are sent to the field-programmable gate array (FPGA) I/O to control all linear actuators for execution. The FPGA is compiled to handle all I/O of the CompactRIO and provide proportional-integral (PI) control of the heater (used to maintain the heart simulator housing temperature at 37°C (body temperature)).