Design of a Mock Circulatory Flow Loop

Jared M. Bieniek

Artificial Heart Laboratory, Department of Bioengineering, The Pennsylvania State University

Sponsor: Keefe Manning, PhD

EXECUTIVE SUMMARY

Heart valve disease leads to the 225,000 heart valve replacement surgeries each year (1).  In vitro studies have shown that mechanical replacement heart valves cavitate in solution.  In order to study stable bubble formation from mechanical heart valves in a physiological setup, a mock circulatory flow loop had to be designed and built prior to experiments.  The design included a left ventricular chamber with adjoining mitral and aortic sinuses.  Pulsatile flow was generated from a piston pump.

Pressures were recorded in the ventricular chamber and the compliance chamber.  Analysis of the loop pressure waveforms was performed to determine the similarity between the mock loop and a physiological cardiovascular loop.  During the pressure waveform analysis. a 120/80mmHg systemic pressure curve was attained with a cardiac output of 5.0 L/min.  The aortic pressure trace appeared similar to physiological curves.  However, problems were encountered with oscillations of the left ventricular pressure transducer, though future work will likely correct for this error.  The mock loop has now been proven to operate physiologically and can be used for other future experimental studies.

INTRODUCTION

Mechanical heart valves (MHVs) are used in many of the annual 225,000 heart valve replacement surgeries (1).  Biological flowing fluids, blood in particular, are subjected to high changes in flow velocities due to the pulsatile nature of biological flow resulting in the possible formation of vaporous cavities.  The process of cavity formation in response to a local pressure drop is known as hydrodynamic cavitation (2).  Cavitation near a MHV can be seen in the image to the right.  Most of the vapor bubbles that form immediately diffuse back into solution, but any vapor bubbles that remain in the low pressure environment have the potential to diffuse and collect gas from the fluid and form stable bubbles.  These bubbles can then travel through the systemic circulation as emboli and can cause blood flow blockages in small cerebral capillaries and lead to strokes.  Microemboli formation by the MHV as a result of cavitation, another problem with MHVs, has been detected in numerous in vitro studies (3-7).  Microemboli have been found in the blood flow of heart valve replacement patients using ultrasound techniques (8).

This design project was performed in order to study the formation of stable bubbles following cavitation from mechanical heart valves in a physiological setup.  Therefore, it was required that a mock circulatory flow loop be constructed to match natural blood flow conditions as closely as possible.

OBJECTIVES

The following objectives were set as criteria for the mock loop design:

• Mimic physiological blood flow conditions
• Contain no fluid leaks
• Easy to empty and fill
• Ability to take pieces apart for cleaning and other purposes
• Include downstream resistance and compliance elements
• Include downstream mock venous reservoir
• Optical access for imaging
• Create physiological left ventricular pressure waveform within ventricular chamber (120mmHg peak pressure)
• Create physiological systemic pressure waveform (120/80mmHg)

The following figure represents the physiological left heart pressures strived for within the mock loop (9).

CONCEPTUAL DESIGN

Two possible design solutions were created for the mock circulatory loop.  Each can be viewed below (top image is design one and bottom image is design 2).

Mock circulatory flow loop design one

Mock circulatory flow loop design two

The mock loop design one (top) contained some discrepancies from a physiological flow loop.  The major difference was that it contained no true model of the left ventricle and it separated the mitral and aortic valves.  The second loop design (bottom) was created to correct for these problems.  By creating a ventricular chamber and placing the valves at angled positions on the single chamber, a more natural setup was attained.  The second design was ultimately approved and detailed designs of the individual acrylic components were generated.

DETAILED DESIGN

Four pieces of the mock circulatory flow loop were custom designed using AutoCAD^® 2004 (Autodesk, San Rafael, CA, USA).

1. Atrial chamber
2. Aortic chamber
3. Pump elbow
4. Ventricular chamber

All four acrylic pieces were designed to connect to one another for the experimental setup.  In the image of the operational loop below, the ventricular chamber is indicated.  The 90˚ angle connecting it to the piston pump is the pump elbow.  The acrylic piece coming off the ventricular chamber at an angle is the aortic chamber and the piece below it parallel to the table is the atrial chamber.  Mechanical heart valves are placed at the joints between the ventricular chamber and the atrial and aortic chambers.  In this way, flow is directed into the ventricular chamber through the atrial chamber when the pump retracts and out the aortic chamber when the pump pushes.  The compliance chamber and reservoir are also indicated in the image.  A hose clamp is placed where the resistance is indicated to restrict the flow.  Pressure is monitored systemically through a bleed tube in the compliance chamber.  The pressure transducer within the ventricular chamber can be seen coming out of the top plate.  Control of the piston pump comes from the combination of the PC and pump controller boxes next to the computer.

The experimental setup image fails to depict some of the other data analysis tools used.  To monitor the flow, a flow meter was placed directly over the Tygon tubing proximal to the compliance chamber.  Signals from the pressure transducers and flow meter were monitored on an oscilloscope and recorded directly to another computer using a data acquisition system.

DESIGN PERFORMANCE EVALUATION

A pressure waveform analysis was performed to assess the degree to which the mock loop mimicked physiological conditions.  Initial results showed that the aortic (systemic) pressure waveform looked good, but there were severe issues with the left ventricular (LV) waveform.  Upon analysis of the loop, it was noticed that the ventricular chamber pressure transducer was oscillating with the flow within the chamber.  It was hypothesized that these oscillations were the cause of the high frequency noise observed.  To correct for the error, the LV pressure signal was passed through a 10Hz low pass filter.  After this alteration, the waveform looked better, though still not ideal.  The pressure traces were recorded and can be seen for two seconds in the plot below.  Flow was successfully maintained near 5.0 L/min for the test trials.

The following table displays the overall results of the pressure waveform analysis versus the desired results.  All parameters were adequately met once the LV signal was filtered to remove much of the high frequency noise.  The LV peak pressure is still somewhat lower than desired and the waveform not physiological, but with the removal of its free movement it is believed that these factors will be corrected.  Future work will focus on eliminating this aberration and moving onto future experiments examining stable bubble formation within the mock loop.