Schematics, key components, and previous iterations can be found here.
Figure 1: The schematic of the mock flow loop shows all critical components, flow direction, and expected flow pressures.
The final design of this project was a mock flow loop to measure the impact of aortic grafts on arterial pressure in the aorta. Standard sized Tygon tubing was utilized to connect the components within the flow loop. Within the circulatory system, flow is provided by a pulsatile pump that outputs a typical human pulse waveform with a controlled flow rate of 60 ml/s and 60 beats per minute. Flow to the model aorta passes through a compliance chamber and then to a ½" to 1¾" tube adapter. From the aorta, flow splits into two directions (as indicated by the dark red arrows in the above Figure). In one direction, flow exits through four aortic arteries, each passing through a tee fitting and to a globe valve. In another direction, flow descends down the aorta, passing by the two gaskets, through one port of the tee tubes. Both flows return to a reservoir that feeds back into the pump.
Figure 2: Annotated photo of the final design setup and in operation
The mock flow loop, once assembled and sufficiently filled with water, was able to read 120 mmHg Systolic Blood Pressure and 80 mmHg Diastolic Blood Pressure. This ΔP of 40 mmHg was a function of the ViVitro pump’s stroke volume as well as the water level inside the compliance chamber. Once this ΔP was obtained, valves had to be shut some amount such that the pressure reached the desired threshold (analytical approach was taken using Bernoulli's equation). It was found that the branching artery valves had to be anywhere from 70-90% closed to reach physiologically accurate flow rates, while the descending aorta valve had to be about 50% closed. This distribution was verified by calculating the flow rates of each artery by measuring the amount of fluid that fills up a laboratory flask over 20 seconds.
A rigid model of an aorta afflicted with a thoracic aneurysm was central to the design of the mock flow loop, and its design was a chief driver in many final design decisions. It was decided that a custom 3D-printed model of an aorta from a patient-donated CT (computed tomography) scan was a practical and affordable solution to meet all the functional requirements and achieve the project objectives.
The functional requirements of a 3D model of an aorta:
Afflicted with a thoracic aneurysm
Life-sized
Anatomically accurate
Physiologically sound model (pressure, flow direction, flow rate)
Airtight
Easily integrate with the flow loop
Note: For (2), (3), and (4), the requirements were not explicitly quantified. And although a range of values can be found published in medical journals, the aorta models that were provided by the sponsors were pre-approved by Dr. Barleben (footnote here).
Figure 3: 3D-printed model aorta using a Stratasys Connex3 SLA printer. Material chosen was VeroClear.
The tube adapter is a component that allows us to utilize the final model aorta’s barbed inlets and outlets. A method to connect the model aorta in series with the pulsatile flow of the pump was necessary to complete the mock circulatory loop. After a couple iterations, standard tubing and fittings were chosen to seal around the barbs of the model aorta. Tube adapters that reduce the cross-sectional area of the flow are inexpensive and allow greater replicability of this mock circulatory loop.
The Tube Adapter consists of a threaded connection between tubing adapters, as seen in Figure X. Tubing connected to the model aorta is to wrap around the larger diameter barbed side so as to provide a water-tight seal. This then acts as a reducer, allowing connections to various smaller diameter tubing to ensure compatibility with the rest of the mock flow loop. This has the benefit of being standard in sizing, thus more commercially available. It also makes use of the barbs on the final design of the model aorta. Tubing should be able to be stretched over both the barbed inlet and outlet of the model and the two ends of the Tube Adapter, and, if necessary, hose clamps and silicon sealant may be applied to ensure a water-tight connection.
Functional Requirement of an Aorta-Tubing Adapter:
Water-tight seal between aorta and tubing
Connect two different-sized openings
Figure 4: Ascending aorta Tube Adapter
Figure 5: Ascending aorta Tube Adapter
Figure 6: Descending aorta tube adapter
Key to quantifying the success of the flow loop for testing TEVAR surgery is a way to measure pressure along aortic arch branching arteries. If at any point during the surgery, blood pressure to any of these arteries that flow blood to the brain drops below a tolerable limit even momentarily, the patient can experience a potentially fatal stroke. Thus, a pressure sensor setup takes this flow loop from being a useful tool for practice to a verifiable, scientific instrument.
Fluid directly interfaces with the sensor barbed port. To implement this setup, the line must be filled with fluid prior to connection with the sensor and then the barb put in a completely full line, otherwise air left in the line would cause artificial pressure.
Figure 7: Circuit diagram of the final pressure sensing system
Figure 8: Soldered filter cap
Figure 9: Overhead view of filter cap with wiring to sensors
Figure 10: Schematic of pressure sensing system
Figure 11: Complete wiring to all 4 sensors
Functional Requirements of Pressure Sensor Setup:
Must be able to sample pressure fast enough to notice if there would be a fatal drop in pressure
The average human heart rate is between 60-100 bpm and about 100 samples in a beat are needed to accurately capture the shape of the beat. Therefore, the monitor needs to capture data at a rate of at least 167 Hz.
Pressure data must be easy to upload and analyze during or after surgery to see success
Pressure data must be measured accurately from the three aortic arch branching arteries simultaneously
A dual-seal method using the combination of two valves: one to seal with nothing inserted, and another that sealed around the catheter. These valves would be held by a 3D printed gasket. The total cost for two gaskets and valves is about a hundred dollars. The reason this design was chosen is because it presented the most reasonable chance for success in meeting the functional requirements while also being the most inexpensive option. Although all options have known feasibility as proven by other transcatheter in-vitro flow models in existence, the silicone seal had much higher risk due to single-use silicone molds and silicon’s tendency to deteriorate. Furthermore, the insertion catheters, although saving design time, much exceed our budget. This double seal model has been used in other concepts of in-vitro flow loop designs and was accepted as a usable method to maintain hemostasis during catheter insertion.
The valves themselves were ordered from Minivalve, a valve producing company in the Netherlands. The valve-holding gasket was designed using Solidworks and 3D printed on a Stratasys J750 with VeroClear: a high hardness and semi-translucent material. The design was printed with the intent to include bolts to compress and seal the valves to the gasket.
Functional Requirements of Catheter Insertion Point:
Proper sizing to allow for insertion of catheter
Be able to introduce two catheters simultaneously.
Create seal between catheter and gasket (no loss of pressure)
Figure 12: CAD model of the Catheter Insertion Point gasket
Figure 13: The two silicone seals compressed by the gasket
Figure 14: First iteration of mock flow loop that was able to record pressure.
Pump Outlet
Ascending Flexible Barbed Cuff
3D-printed Barbed Model Aorta
Branching artery T-connector
Globe Valve
Tubing connecting T-connector to pressure sensor
Reservoir
Abdominal Flexible Barbed Cuff
Pump Inlet
The final design looks similar to the above Figure 14 of the first iteration of the mock flow loop; however, the Aorta-Tubing Adapter design changed from a 3D-printed Flexible Barbed Cuff to a Tube Adapter that takes advantage of standard tubing sizing. Starting with the pump outlet (1) flow is pushed toward the model aorta (3). To connect tubing to the entrance of the model aorta, there needs to be some Aorta-Tubing Adapter that allows connection of inlet to the aorta. This was done using the Flexible Barbed Cuff (2) in the imaged setup. At steady-state, the flow is sent through the model aorta and exits at the abdominal region of the model where another Flexible Barbed Cuff (8) can be found and each of the branching arteries connected to barbed T-connectors (4) that keep the same diameter of the tubing they direct flow from, but also has tubing (6) that allows connection to the barb of the pressure sensor. The blood-substitute fluid then goes through the globe valves (5) from each branching artery, which then terminates in the reservoir (7). Similarly, the flow from the model aorta through the abdominal Flexible Barbed Cuff (8) passes through a globe valve and also terminates in the reservoir (7). The pump then sucks the fluid out of the reservoir at (9) sending the fluid back to the start (1). Below are descriptions of each of these components. Experimenting with this setup, it was clear that there were a couple of problems:
a. Inaccurate pressure readings: Negative pressure readings meant the pressure sensor was at risk of being damaged, and there should not be negative pressure in the tubing of the pressure sensor
b. Leaking: leaking at point (8) in the Figure 1 above resulted from improper tubing connections and incompatibility between the Flexible Barbed cuff and the 3D-printed model aorta (3)