ULTRASOUND-INDUCED THERMAL THERAPY OF HYPERPLASIA IN RINGED EXPANDED POLYTETRAFLUOROETHYLENE (ePTFE) ACCESS GRAFTS
Kidney dialysis access grafts can be rendered useless or dangerous from scar-tissue-like cell growth. Non-invasive mild heating of these grafts with ultrasound prevents/reduces harmful growth in lab environments and simulations.
After more revisions than I can remember, the master's thesis is done.
Literary critics have called it
- my wife
"Captivating, a page-turner. Couldn't keep it out of my mouth."
- my 1-year-old
If you have an insatiable taste for heavy, scientific reading, find it here.
The research was actually an enjoyable process. Tons of math, stats and programming. The killer was the English revisions. Shout out to all those English-As-A-Second language folks that walk the gauntlet of grammar and content revisions.
A Bit More Detail
My research was broken into three phases.
- Cell Exposure Experiments
- Ultrasound and Heat Transfer Simulations
- Transducer Design
The cell exposure was designed to grow SMCs (smooth muscle cells) and ECs (endothelial cells) to observe their response to mild temperature increase on the graft material (ePTFE, expanded polytetrafloraethylene, essentially porous teflon). SMCs make up the undesirable growth. ECs represent good healthy surrounding tissue cells. Turns out SMCs are more sensitive to mild hyperthermia than ECs. Good News!
Ultrasound and Heat Transfer Simulations and Lab Validation
This is the crux of the idea. Teflon (ePTFE) is very absorptive to ultrasound energy. The hope was that we could generate selective heating in the graft material and not so much in the surrounding tissue.
The simulations are completed in two stages:
- Simulate the propagation of the ultrasonic beam and find the deposited ultrasonic power
- Simulate a temperature rise in the media of interest based on the deposited ultrasonic power
The lab validation involved:
- Measuring material properties (of very thin samples). I documented a method not yet in the literature for this (although not mind-blowingly novel)
- Building tissue simulate model, heating with ultrasound, and measuring the heating
- Comparing to simulations
Ultrasound therapy, in contrast to imaging, requires the ultrasound energy to remain active for long periods (seconds) compared to imaging (millionths of seconds). The simulation is not concerned with the transient ultrasound, or the energy that dissipates so quickly as to not deposit appreciable power. The simulations that disregard these transient effects are called steady-state.
There are a number of methods available for steady-state simulations, but in the case of an inhomogeneous medium (where the mechanical parameters of the medium are not the same across the whole area of interest) there is a method called Hybrid Angular Spectrum (HAS). The method was developed by Douglas Christensen at the University of Utah, who just happened to be my thesis advisor (lucky me!).
HAS is worth a discussion for anyone that wants to talk about developing a commercial ultrasound simulation tool that, in its niche, dominates currently available commercial methods. HAS runs on a regularly spaced grid (more discussion later).
Back to the point at hand, we simulate the ultrasound energy into the graft and surrounding tissue and then based on known tissue and material properties determine the absorbed power.
Heat Transfer Simulations
Heat Transfer is a bit more plug and play as far as we are concerned. Take the deposited power, put it into an FEM (finite element method) solver (COMSOL) and simulate the ultrasound being on for various times and the associated temperature rise.
The tricky part of is getting from the HAS regularly spaced grid output to the COMSOL FEM mesh. I put together an algorithm to do this. One of the nice benefits is the ability to get from standard 3D drafting programs to HAS grids. Again another shameless plug for commercializing the process. It isn't that far of a jump.
The simulations predicted selective heating in the graft! Good news!
Oversimplifying here, but essentially stick a graft in some tissue simulate material. Install some thermocouples to monitor temperature. Hit it with ultrasound and watch the temperature.
Lab validation showed correlation to the simulation results! Good news again!
Design and Simulate a Hybrid Therapeutic and Diagnostic Transducer
Design a transducer that has a space carved out for an "off-the-shelf" linear array for imaging during treatment. Simulate the result of carving out the imaging array.
Given continued funding, it makes sense to move to animal studies. Researchers should be wary of how long the grafts have been installed because their mechanical properties change based on how much cell growth has become infused in the graft material.