 Our next presenter is Chris Smith. He's a fourth year medical student as well at the University of Texas Medical Branch. Thank you. My name is Christopher Smith. I am a medical student at the University of Texas Medical Branch. As you said, I live in Austin, Texas, and it's been a great opportunity to come to the Moran Eye Center and work with you. Today I want to present on a project I worked on looking at the confocal microscope, comparing two different lenses on there among DSEC patients. To start off, I'd like to review the anatomy since it is relevant to my project of the cornea. These are cross-sections, stained cross-sections of the cornea. On the top left, we'll mainly focus here. We have the layers, the epithelium, the stroma, and the endothelium. I was mainly focused on the corneal endothelium here that's the most posterior layer and is in contact with the aqueous humor. Since the cornea is in a vascular layer, it is important to get nutrients. This endothelium is important to get nutrients into the stroma. It does this by a pump leak mechanism, meaning it leaks the solutes in, but it needs to get the water out so that the cornea does not become over-hydrated and deflect light. An example of this is seen in fuchs dystrophy, where gutata are built up in the cornea and over time the corneal endothelium thins and the cell layer is lost and so it's lost function. When we lose the function, the cornea becomes more cloudy and less clear, so light is not able to go through it. There's different ways of correcting this by surgery and with different transplants. The corneal endothelium, one way to calculate the parameters or the healthiness of this graft is by doing an endothelial cell density. The way they do this is they take a picture with a microscope of the endothelium and you literally take the area of the, you count a number of cells in a contiguous area and divide that over a certain number of area. This gives you a density. This is, pertinent, at birth we have the most, the highest density, around 4,200 cells per millimeter squared and then over time you can see that the density drops and in adult, during our adult years it plateaus to around 2,700 and then we lose 0.6 percent per year and with, obviously with pathology or diseases, we lose even more. They have been looking at this ECD. One of the projects of this cornea donor study, one of their goals was to see how corneal endothelium, this ECD, pertains as far as follow-up in the clinic and how useful of a parameter it is. This study that performed in 2010 looked at about 500 graphs. 17 of these graphs had failed and 483 of them were healthy after transplant, corneal transplant and they went back and looked at the times of, at different times, oh sorry, at different times of the endothelial cell density and were able to measure is this a useful parameter. So to look at some of the figures of these studies we see on the top, this is our healthy graphs and below the solid lines are graphs that had failed and as you can see the endothelial cell count is much less. Looking further of significance of this is they broke it into three different groups. At six months they saw a group with 17, less than 1700 and then these other two groups with greater than 1700 and as we can see the percent incidence of failure is much more in this solid group. Thus they concluded that at about six months this is a good follow-up as far as looking at a parameter to follow up these either desicc or corneal transplant patients to see how healthy their graft is. This is the confocal microscope. It is on the fourth floor of the morocular brand eye center. If you haven't been able to see it you also take a field trip. It's a great machine and we spent lots of money fixing it and it's working great now. There has been over time a couple different techniques of visualizing the corneal endothelium that have been used in the past. One specular microscopy has been very popular as well as confocal microscopy. Specular, this is the first image taken in published in 1974 of the corneal endothelium. You'll see these hexagon cells and this is the one cell layer thick and about ten years later that's when the confocal microscope started being more wide used to visualize this parameter and calculate the ECD. The difference between these two techniques can be seen here. The confocal has two different pinholes that you look through. One is the illumination where the light is coming from and it focuses on a specific layer right here and bounces off and goes through the observer or the recorder through this pinhole. It's a good useful tool of the confocal microscope since the illumination and the observer are focused on one specific layer. Any light out of focus or deeper is not able to reach the observer. Thus any other light through different layers that can blur the image is not reached to the observer and we get a much clearer crisper image as opposed to the specular microscope which is a lot more simple of bouncing off a certain layer of the light and into the observer and with the different angles the light that is most direct as it bounces off is the brightest light and anything that gets deflected you get these dark. This would be an example this is a cartoon of a gutata that is not shown up on the picture. It's a little more dark. There have been papers that have compared these two techniques. This is a very important paper to the project I worked on since I sort of modeled off what they did and kind of took it a few steps further. This is the abstract and the main things I highlighted from this I wanted to look at. They looked at three different techniques. Two lenses on the confocal microscope as well as specular microscope and compared they took several images of desec eyes and 20 normal eyes and they had a masked observer after they took these images to just label these excellent, good, fair or poor. Poor would be considered uncountable so they weren't able to get an ECD at all. Their conclusions found that the image quality was best with the contact confocal microscope lens and then the non-contact and then the specular microscope in that order. We can look at their figures. Here is an example of what these images look like. Here excellent, good, fair, poor. This is an example of what these images would look like here. Poor as you can see can't pick out cell borders here very difficult to determine an endothelial cell density and it gets a little better as time and compared to the confocal nice, crisp and clear cell borders is easy for the technician or whoever is counting these things to put a dot in there and get an accurate endothelial cell density. They broke them down into percentages and comparing the confocal versus specular they found it was very significant of how many were failed or were not able to get a calculable endothelial cell density. This was 42% of the images that they looked at. They did not have any failures here in the non-contact and this is where I branched my project. I wanted to compare these two lenses because as I started examining these patients I noticed that it was a little more difficult to get the non-contact confocal microscope. Sometimes it took several minutes to get an image even clear enough to be countable. Thus I wanted to calculate what I was seeing as far as is there a failure rate here? Is this useful as far as clinical efficiency? Because if you talk to many of the techs, their goal is 12 minutes to check in the patient, examine, this is including getting a refraction, getting all the different parameters they need to get ready for it when the physician comes in. If they have to spend 6, 7, 8 minutes getting a clear image here at this lens type that could be significant and slow the clinic efficiency down. These are among DSEC patients. This would be just quick among normal patients. We can see the images overall are a little better but we get similar results here. Here are an example of the two lens types that are available with this confocal upstairs. There's a 40x and a 20x lens. This is a contact and a non-contact lens that they named as such because this one moves forward. You put a little bit of gel, a drop of gel on the tip of this lens and as you move in it goes and touches, actually contacts the center of the cornea. This is theorized that it will give a better medium for which light can go through. However the non-contact one does not have to touch the eye so the advantages of this would be that the patient would be more comfortable since they don't have to have their eye touched. The purpose of my study briefly stated here is to compare the efficacy of these two contact and non-contact modes in evaluating the endothelium after these DSEC patients. Here are my methods that I used. I took 23 eyes after DSEC and 22 normal eyes. I sat them down in front of the confocal, got them all ready for the exam and before I moved the joystick, that's the first movement towards studying. That's where there's variability in the time of getting this measurement. I put a stopwatch and I estimated about five minutes would be a reasonable time to get a clear image for one of these people. So a clear image and so I started the stopwatch and that's when I started measuring and as you bring the joystick in further and back you want to hit the brightest corneal image as you focus right on the center of that pupil and once you get that bright image the machine will give you audio feedback and you push scan and then it will automatically align and you'll get an image that you can upload to the computer. The contact is similar but you put a drop of that gel and you go in and you touch the cornea. You give them a drop of anesthesia before that of course. Anesthetic. Then after that was done we uploaded, I'll show in the next slide how this is done. We calculated by a variable frame technique. You upload the images and can count them. Then after this the failure rate was determined by the inability of the two lenses to obtain readable image within five minutes. Since there has been an idea that contact versus non-contact, somebody might not like having their eye touch during the contact mode, I evaluated with just a simple scale of discomfort. I asked the patients on a scale from one to five, one being no discomfort at all and five being the maximum discomfort they can experience. What was their experience as far as these measurements after both were performed? So after we get a picture, an image, we bring it up to the software and this is a software program. This is an example of an automatic software. You can push these images. You push this button right here and it comes up and the software can find, by contrast, it can find these cell borders and quickly get a rough endothelial cell density. However, they have found that this is not very accurate since, as you can see here, it's picking out cells that maybe aren't there or missing cells that not splitting two cells that are separate from each other. This is the variable frame technique that I used in my project, a more manual count. What you do is after you upload, you count these number of cells, you put a dot about in the center and then you take this blue tool and you outline all the way around and after you get all the way around, it will automatically tell you the number of cells you had and the area and divide it and give you an endothelial cell density as seen here below. This is an example here. I kind of wanted to show the image quality that I was seeing between these patients. This is a desec patient. As you upload, this is the kind of image you're going to be counting which is significant because it might be a little difficult to count these images if they are blurry. This is an idea of a contact. I want to compare this one to the non-contact of the exact same patient. As you can see, a little more difficult in these non-contact images, this would be considered a countable image in my study. This was typically what I saw. However, the borders are a little more, is this a cell here? Is this one cell a little more difficult to determine and can maybe influence our decision as far as the follow-up of these patients? This is my eye. It looks pretty good, right? I hope. I just wanted to show an example of a normal patient to compare between the desec eyes. This would be a contact in my eye and, once again, my wonderful eye here. The results I found, I concluded where neither method failed among the normal eyes, the contact or the non-contact lens. I was able to get a countable image in each one of these. Significantly though, in the desec eyes, the failure rate was 26.1% in the non-contact mode and none of them failed in the contact mode. So it was much easier to obtain an image within five minutes using this contact mode. The endothelial cell density while comparing these in between patients what they counted, it counted about on average 61.9 cells more dense in the contact method which really isn't clinically significant and there have been several papers that you can read on this in the past that they have compared these two modes as far as calculating the density and they come up with very similar results once you get an image. However, if there is a failure, this could be significant. This could slow you down. And then looking further, the failure of non-contact mode was not correlated with best corrected visual acuity, central cornea thickness, age or length of time from surgery. The patients reported comparing the discomfort, there was no significant difference between these two methods. And this is just a figure I made for this showing that there is no difference between contact and non-contact mode. They are around 1.7 as far as what they come to remember. One being no discomfort at all, five being maximum discomfort. My conclusion that I found the contact lens on the CONFLOSCAN 4 was a little quicker and it got the clearest image. When the non-contact mode works lens, it can be quicker. However, if you have to redo the lens because you didn't get a clear image the first time, it was a little more difficult to use. And then as I say before, the mean ECD was a little higher but not significant. And then the subjects did not prefer one method over the other. I'd like to give acknowledgments to Dr. Mark Mifflin and Dr. Maylon Xu who helped me out with this project. And are there any questions I could try to answer, please? Exactly. The calculating it manually, that was not put into the five minutes. That took me approximately not too long, maybe another five to ten minutes, something like that. Maybe even shorter. I'm not exactly sure. I never timed myself, but I got quicker as I had to do several of these, obviously. Exactly. Which either way you're going to have, which is interesting because in the non-contact mode, if you have to measure this with the variable frame technique, as I said, if you can't see those cell borders, it was a little difficult. It did take more time. And if the tech is in a hurry, it'd be easy for them to just flub through the borders and, you know, this might be border. I don't really care. I got to get going. So that could be another thing to look at. Yeah. And further, I forgot to mention with the six months, they noticed that at six months that was significant. However, looking, they found that there were some endothelial cell densities in the corneal transplant paper that got lower than 500, yet they still had clear graphs, which is something that Dr. Mifflin has seen even with his patients. So it is more of an over time following as opposed to a specific number that you reach because even low numbers can keep the corneal clean. That's true. All right. Thank you.