Daniel Carlin's Past Research Projects

Here are the previous research project that I have been involved in:

Project Title Location Lab 
Gene therapy for collapsed intervertebral disks University of Pittsburgh Medical Center Ferguson Lab Summary
In silico cardiac model for ultrasound research Duke University Trahey Lab Summary
Diffusion simulation for orthopedic scaffold design NC State University Haider Lab Summary

Gene therapy for collapsed intervertebral disks

My first research experience began in the summer of 2004, and continued into the summer of 2005. I was given a position in the Pittsburgh Tissue Engineering Initiative as a research intern in the Ferguson Laboratory for Orthopedic Research under the direction of Dr. Lars Gilbertson and Dr. James Kang. The focus of this particular project was regeneration of intervertebral disks. Collapsed disks are the leading cause of back pain, and the current cure is a fusion of the adjacent vertebrae, which is highly invasive.

"Black disks" as they are known because of their appearance on MRI scans, have decreased hydration due to a decrease in glycoaminoglycans (GAG), which are polymers that hold water in tissue due to a fixed partial negative charge that interacts with the highly polar water molecules. The essential problem is that in order to produce GAG, the cells must increase their metabolism above normal levels, and require nutrients and oxygen to do so. Intervertebral disks are avascular tissue, so all oxygen and nutrients are supplied to the cells in the center of the disk by diffusion, and the cells are under constant nutrient stress. If the cells are induced to produce GAG, their nutrient supply must increase. However, if the cells begin to produce more GAG, the diffusion properties of the disks change, and the extra fixed negative charges, which also aid in diffusion, may serve to counterbalance the increased nutrient demand.

To test whether this was the case, we utilized a technique developed by Urban et al. whereby the diffusion stress was simulated by placing disk cells between two glass slides, so that nutrients could only diffuse from the gaps at the edges. After a few weeks, the slides where disassembled and a red-green viability test was performed. Naturally, the cells near the edges of the slides were more viable than those near the center, because the nutrient did not have to diffuse as far, and thus were in greater supply. The test, then, was to use gene therapy to promote production of GAG, and see if the therapy was successful in both promoting GAG production and the viability of the cells.

I undertook this project as a member of a three person team made up of myself, another undergraduate and a medical resident. Our duties included the collection and isolation of bovine and human disk cells, the culturing of all of the cells until they were ready to be transduced, the transduction of the cells, the construction of the glass slide "diffusion chambers," the staining and imaging of the final cells, and the analysis of the data. Initial results were promising, and were presented by Dr. Adam Shimer, our resident, at the 2004 Ortho Mare Nostrum Conference. However, we were unable to consistently reproduce our results, although the protocols that we developed are still in use at the Ferguson Lab.

In silico cardiac model for ultrasound research

My second, and most extensive, research project was conducted under Dr. Gregg Trahey at Duke University, in the Pratt Undergraduate Research program. I began the project in January 2006 and finished in May 2007. The topic of this project was cardiac ultrasound, and my goal was to create a beating heart model that could simulate realistic cardiac ultrasound. This project was almost entirely mine, although I conducted it under the guidance of a graduate student.

Cardiac imaging is a very common use for ultrasound, due to ultrasound's real time capabilities and lack of radiation exposure. Another newer advantage of ultrasound is its ability to track motion by correlating the images from one frame to the next in a process called speckle tracking. However, speckle tracking in the heart is greatly complicated by the natural motion of the heart. The goal of this project, therefore, was to create an in silico model of heart motion for optimization of speckle tracking in cardiac ultrasound imaging.

The idea was to adapt an existing finite element model of the heart, called Continuity 6, for use in conjunction with ultrasound simulation program Field II, which operated in MATLAB. We obtained the 3D coordinates through time of certain locations of tissue of a normally beating canine heart. These locations were the "nodes" of the model, i.e. the corners of the elements. Field II takes, as input, the location of "scatterers" which in real tissue represent cell nulcei or any other material with different sonic properties from the surrounding tissue. The key, then, to this project was to get the motion of the scatterers to match that of the nodes in the simulation. To do this, I performed an interpolation of the motion of the nodes to that of the scatterers. Once we obtained a realistic set of coordinates for the scatterers, we input them into Field II, which I had set up to simulate the Siemens VF 10-5 ultrasound probe. Finally, after we obtained simulated ultrasound images, the motion of the tissue was tracked using the texture of the images. This gave us a good idea of how well the motion estimates we got by watching the ultrasound images matched that of the actual motion of the nodes in the original model.

The results that we obtained in this project were very promising, and we were preparing a paper based on the results. However, a very similar project was published by a group at Columbia in March 2007, so we could not immediately publish the results. However, further investigation and use of my model is in the works under Dr. Trahey.

Diffusion simulation for orthopedic scaffold design

My most recent project was conducted in the NC State department of mathematics, in the REU for Applied and Industrial mathematics. I worked under Dr. Mansoor Haider on articular cartilage scaffolds for repairing defects in knee tissue. The scaffold acts as a medium for new cells to grow in and so that they can repair the defects. The overall goal of this scaffold is to match the mechanical and diffusion properties of the surrounding tissue so that under stress the scaffold gel does not separate from the adjacent tissue, but also allows sufficient nutrient diffusion to newly growing cells. Both the mechanical and diffusion properties of the scaffold are functions of the molecular weight and density of the polymer that makes up the scaffold. Also, particles diffusing in are not uniform in size, so their diffusion properties are affected differently by changing the properties of the scaffolding.

My project, which I undertook with one other undergraduate and one graduate student, was a simulation of diffusion of different sized particles through scaffolding of different molecular weights. The core of the simulation was a "random walk," in which at each time step the particle moves a set amount in a random direction. This is the generally accepted model of diffusion of particles, accounting for bouncing off the particles of the medium (in this case water). In addition to this random walk, however, we added the additional rule that the particles would bounce off the scaffold, which was represented as cylinders of various radii and lengths, depending on the scaffold properties that we wished to represent. For each data point, several thousand of these simulated walks were performed, and the diffusion data was extracted from the final position distributions of the particles.

We found that the diffusion properties were decreased by increased volume taken up by the scaffold, and by increased radii of the particles. This matches qualitatively with experimental results, but the model still needs to be refined to match quantitatively. Likely possibilities for refinement are in the geometry of the scaffold and the simulation of the bouncing of the particles off the scaffold. Dr. Haider has also indicated that he will try to refine the results into a publication.

My Homepage