August 2004
Volume 45, Issue 8
Free
Cornea  |   August 2004
Monitoring of Rabbit Cornea Response to Dehydration Stress by Optical Coherence Tomography
Author Affiliations
  • Kamran Hosseini
    From the Center for Biomedical Engineering, and
    Department of Ophthalmology, University of Texas Medical Branch, Galveston, Texas;
    Department of Ophthalmology, University Hospital of Maastricht, Maastricht, The Netherlands;
  • Alexander I. Kholodnykh
    From the Center for Biomedical Engineering, and
  • Irina Y. Petrova
    From the Center for Biomedical Engineering, and
  • Rinat O. Esenaliev
    From the Center for Biomedical Engineering, and
    Departments of Physiology and Biophysics, and
    Anesthesiology, University of Texas Medical Branch, Galveston, Texas.
  • Fred Hendrikse
    Department of Ophthalmology, University Hospital of Maastricht, Maastricht, The Netherlands;
  • Massoud Motamedi
    From the Center for Biomedical Engineering, and
    Department of Ophthalmology, University of Texas Medical Branch, Galveston, Texas;
Investigative Ophthalmology & Visual Science August 2004, Vol.45, 2555-2562. doi:10.1167/iovs.03-0792
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to Subscribers Only
      Sign In or Create an Account ×
    • Get Citation

      Kamran Hosseini, Alexander I. Kholodnykh, Irina Y. Petrova, Rinat O. Esenaliev, Fred Hendrikse, Massoud Motamedi; Monitoring of Rabbit Cornea Response to Dehydration Stress by Optical Coherence Tomography. Invest. Ophthalmol. Vis. Sci. 2004;45(8):2555-2562. doi: 10.1167/iovs.03-0792.

      Download citation file:


      © 2016 Association for Research in Vision and Ophthalmology.

      ×
  • Supplements
Abstract

purpose. To evaluate the application of optical coherence tomography (OCT) for continuous noninvasive monitoring and quantification of the dynamics of corneal response after exposure of the cornea to dehydrating stress.

methods. The changes in central corneal thickness (CCT) and scattering properties of the cornea were monitored with OCT in rabbit cornea in vivo after topical application of a glycerin-based hypertonic agent (HA) or prolonged surface evaporation of the cornea. The observed changes in backscatter were correlated with the changes in corneal hydration.

results. An inverse relationship was found between the logarithm of the intensity of backscatter within the cornea and the degree of corneal hydration at which the intensity of the backscatter changed up to 20 times between the peak of the de- and rehydration phases. An analytical relationship is derived between the magnitude of the backscatter from the stroma and the extent of corneal hydration. Furthermore, depending on the concentration of the drug, a peak overshoot in corneal thickness in the range of 40% to 90% was detected during the rehydration process after topical application of the HA. At a 100% concentration of HA, the average dehydration rate was 74 μm/min, whereas the average rehydration rate was 12.4 μm/min. The same parameters for surface evaporation were 2.7 and 1.5 μm/min, respectively.

conclusions. OCT may offer a unique capability to quantitatively monitor the dynamics of corneal response and to assess corneal function based on noninvasive detection of the changes in the optical properties and morphology of the cornea after topical application of dehydrating agents.

Characterization of the mechanism of corneal hydration control (CHC) and the development of new techniques for quantification of water content and its distribution within the cornea have been the focus of many studies. 1 2 3 Accurate quantitative assessment of corneal hydration or the degree of stress-induced changes in corneal hydration or corneal thickness could have a significant impact on the diagnosis and treatment of various corneal disorders such as Fuch’s dystrophy, the precision of the “cut” in refractive surgeries, and cornea transplantation procedures. 4 5 6 Furthermore, variation in corneal hydration can affect the accuracy of the measurements by diagnostic tools, such as the clinically used intraocular pressure tonometer and the corneal topographer. 7 8 9 In addition, the effects that drugs may have on corneal hydration can provide valuable information about the function as well as the interaction of drugs with the cornea and the subsequent changes in CHC. 10 11  
Development of an optical technique for noninvasive monitoring of dehydration-induced changes in corneal morphology will help elucidate the subtle effects of topical drugs on corneal morphology. Furthermore, this approach could be extended to assess quantitative corneal responses to stress, thus allowing for functional imaging of the cornea. 
In the past decade, optical coherence tomography (OCT) has gained great acceptance in the field of ophthalmology because of its high-resolution imaging capability. 12 13 14 15 This optical imaging modality produces high-resolution detailed cross-sectional images of tissue in vivo and has proven to be very valuable in other medical fields. 16 17 18  
The unique optical properties of the cornea allow for high-resolution imaging of the entire cornea with OCT. Recent studies have demonstrated the applications of noncontact slit lamp–adapted OCT for in vivo imaging of the human cornea. 15 However, the clinical applications of OCT have primarily focused on high-resolution imaging and qualitative analysis of corneal morphology. 13 14 Recently, we have used OCT for qualitative measurement of the optical properties of the cornea. 19 The strong correlation between corneal thickness and its hydration may promote the use of OCT for quantitative monitoring of corneal hydration in vivo. 
The purpose of this study was to demonstrate the feasibility of applying OCT-based techniques for safe, noninvasive, and noncontact monitoring of the dynamics of rabbit corneal dehydration and rehydration in vivo in real time and for the quantitative assessment of rabbit corneal response to dehydration stress based on the measurements of temporal and spatial changes in optical behavior of the cornea. 
Methods
Instrumentation
An interferometer-based fiber-optic OCT system built at the Institute of Applied Physics of the Russian Academy of Sciences 18 was used in the present study. A superluminescent diode (center wavelength = 1300 nm, incident power on the sample surface ∼200 μW) served as a source of low-coherence radiation for the OCT system. Piezoelectric modulation of the fiber length provided in-depth scanning within 2.5 mm. Electromechanical bending of the fiber tip in the focal plane of the objective lens ensured lateral scanning in the adjustable range of 0 to 3 mm. The axial resolution of this system was approximately 14 μm. In addition to recording OCT images for postprocessing, the intensity of light reflected and backscattered from different regions within the cornea was recorded in logarithmic scale by a digital oscilloscope (model TDS 3012; Tektronix Inc., Beaverton, OR). This allowed real-time measurement of the changes in optical properties and morphologic features of the cornea. Further technical descriptions for this system are published elsewhere. 20 21  
Measurements of Corneal Thickness
Corneal thickness was calculated as the distance between the two major peaks in the oscilloscope signal, which represents reflections from the corneal boundaries (shown with short arrows in Fig. 1 ). In all experiments, the corneal thickness was measured at the center of the cornea with a lateral deviation of ≤0.5 mm. By focusing the OCT probing beam on the center of the corneal curvature, the variance in corneal thickness is minimized. 7  
The thickness of the cornea was measured based on the measurement of the time delay between the peaks. The coefficient of 0.325 μm/μs was used to convert the temporal delay of the oscilloscope measurements into corneal thickness. This coefficient was calculated after measuring the time delay (2.22 ± 0.01 ms) between two peaks of reflection on the inner walls of the optical glass cuvette with a known thickness (1.00 ± 0.01 mm) and use of the average corneal refraction coefficient 22 23 n = 1.38. We assumed a constant value for this coefficient, independent of the hydration level of the cornea within the range of our corneal hydration manipulation. The precision of the relative peak position reading on the oscilloscope display was 8 μs (3.6 μm for air and 2.6 μm for the cornea); therefore, the typical thickness of rabbit cornea (≈350 μm) was measured with an accuracy of better than 99%. Motion artifacts did not affect the accuracy of thickness measurements because a very short time is required to record a single in-depth scan from the cornea (<0.02 seconds). 
Measurements of the Backscatter
Quantification of hydration-induced changes in the scattering properties of the cornea based on OCT measurements requires consideration of several factors that could significantly improve the accuracy of these measurements. The OCT signal produced by scattering in the cornea is 102 to 103 times smaller than that produced by reflection from the corneal boundaries. This is the reason for using logarithmic amplification of the OCT signal. Moreover, there are speckles in the OCT signal that are a random modulation of the amplitude of the backscatter due to the coherent nature of radiation and interferometric design of systems such as OCT. 24 25 Recently, we reported that by performing a square pattern scan with the probing beam on the corneal surface (100 × 100 μm) and averaging the recorded scans, one can minimize the speckle and electronic noise up to level of a few percents of the OCT signals. 19 Once we reduced the noise in OCT measurements, we were able to quantitatively study the spatial scattering profile within the cornea as a function of time. To prevent artificial alteration of scattering profiles from the changes in the OCT in-depth sensitivity, 19 we focused the OCT probing beam on the epithelial surface of the cornea for all our experiments. 
Animal Protocol
For this study we used white New Zealand rabbits, because their eyes are similar in anatomy and physiology to the human eye. All animal care and treatment procedures adhered to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. The rabbits were anesthetized with a mixture of xylazine (5 mg/kg) and ketamine (25 mg/kg) injected intramuscularly as needed. The animals were positioned and stabilized, and their eyes were brought into proptosis and held in place with a rubber ring mounted in a fixed, larger metal ring. This holder was designed to position the eye relative to the OCT probe. The ring helped immobilize the eye while at the same time permitting perfusion of the eyeball tissue. An artificial tear was applied as needed to prevent corneal drying before the administration of the stress agent. This approach enabled us to maintain the corneal thickness close to its initial value with a measured variability of approximately 5%. Ten rabbits were used in the study. One eye of each of four rabbits was subjected to a 100% hypertonic solution (Ophthalgan; Wyeth, St. Davids, PA; more than 98% glycerin, 1.0% water, 0.55% chlorobutanol) as the dehydrating agent. Another two eyes (two rabbits) were subjected to a 50% concentration of the same hypertonic solution to investigate the effects of drug concentration on corneal response. Four eyes from a different group of four rabbits were subjected to evaporation at room temperature as another means of dehydration. 
Changing the Hydration State of the Cornea
Before the hydration manipulations, baseline OCT measurements were performed for 10 to 15 minutes. We then altered the hydration state of the cornea in vivo by topical administration of 1 drop of the hypertonic agent. The strong dehydrating effect of this hypertonic agent caused an almost instantaneous de-swelling of the cornea. 26 The dehydration phase was followed continuously by OCT measurements of the central corneal thickness (CCT) and corneal backscatter. The de-swelling process continued for a couple of minutes and then reached its peak. At that time, the cornea started to rehydrate. As soon as the cornea started to regain some of its thickness and absorb more water, we released the eye from the rubber ring holder. Once the animal’s eyelids covered the eye, we taped the eyelids lightly to ensure a relaxed state for the cornea to rehydrate. Then, we opened and examined the eye every 8 to 10 minutes to record the OCT image and oscilloscope signal and to monitor the changes in corneal thickness and the intensity of the backscatter, respectively. The eye was closed again to allow for the rehydration phase to continue. When the rehydration rate was decreased substantially compared to the initial rate of rehydration (<25%), we reapplied the hypertonic agent to investigate whether the fact that the cornea that had already been treated with the hypertonic agent would cause any change in its response to the dehydrating agent. 
Comparing Dynamics of Different Dehydration
To compare the influence of different dehydration procedures on the dynamics of the hydration cycles, we performed a similar series of in vivo experiments in which the eye of the rabbit was kept open for an extended period of time to allow for slow dehydration of the cornea at room temperature. Continuous OCT monitoring and oscilloscopic measurements were recorded during the free convective surface evaporation phase of the cornea. Once the corneal dehydration rate diminished substantially, we allowed the cornea to rehydrate in the same manner as with the hypertonic agent–treated rabbits. However, it should be noted that during the rehydration phase of these experiments OCT monitoring was not performed continuously, as was the case in the rabbit treated with the dehydrating agent. At the end of each experiment, the animals recovered fully with no complications. 
Data Analysis
To describe the dynamics of the de- and rehydration processes during the stress test, we applied an exponential fit to our experimental data. This fit has been commonly used to describe the corneal thickness recovery from a swollen state. 27 28 29 30 31 In our experiments, CCT changes take place not only from a swollen state but also from a thinned state, and can reach a new quasi steady state with a different value than the original base thickness. To explain these events, we used an exponential fit in two different forms so that both re- and dehydration processes are addressed independently. 
For the rehydration phase after strong dehydration of the cornea, we used equation 1a :  
\[L(\mathrm{t})\ {=}\ (L_{0}\ {-}\ {\Delta}L_{0})\ {+}\ ({\Delta}L_{0}\ {+}\ {\Delta}L_{\mathrm{overshoot}}){[}1\ {-}\ \mathrm{exp}({-}t/T_{\mathrm{hyd}}){]}\]
where (L 0 − ΔL 0) is the corneal thickness reduced from the base thickness (L 0) at t = 0 and ΔL overshoot is the change in corneal thickness due to overswelling of the cornea (compared with the base thickness L 0) that can be achieved, assuming that the process can reach a quasi steady state. It should be noted that time t can start from any moment during the process that we choose to apply the exponential fit for the experimental data. The parameters of this exponential fit are the time constant T hyd and magnitude of the CCT change (ΔL 0 + ΔL overshoot). These parameters were calculated through computer software (Origin 7.0 software; Origin Laboratory Corp., Northampton, MA) by curve-fitting the experimental results. 
For the dehydration phase from the swollen state we used equation 1b :  
\[L(t)\ {=}\ (L_{0}\ {+}\ {\Delta}L_{\mathrm{swel}})\ {-}\ ({\Delta}L_{\mathrm{swel}})\ {+}\ {\Delta}L_{\mathrm{thin}}){[}1\ {-}\ \mathrm{exp}({-}t/T_{\mathrm{deh}}){]}\]
where (L 0 + ΔL swel) is the initial thickness (t = 0), swelling from the base thickness L 0, and ΔL thin is the magnitude of the decrease in corneal thickness from the base thickness L 0. The calculated parameters are T deh and (ΔL swel + ΔL thin). (Note: ΔL thin is a calculated value in the dehydration process, whereas ΔL 0 is a measured value in the rehydration process.) 
The initial rate (dL/dt) at time t = 0 (i.e., maximum due to the exponential nature of the processes) of the CCT alterations can be calculated according to the following formulas:  
\[dL/dt\ {=}\ ({\Delta}L_{0}\ {+}\ {\Delta}L_{\mathrm{overshoot}})/T_{\mathrm{hyd}}\]
 
\[dL/dt\ {=}\ ({\Delta}L_{\mathrm{swel}}\ {+}\ {\Delta}L_{\mathrm{thin}})/T_{\mathrm{deh}}.\]
 
To relate drug-induced changes in the corneal hydration, we used a well-established analytical relationship between corneal thickness and its hydration 32 :  
\[H\ {=}\ a\ {\times}\ L\ {-}\ b\]
where L is the corneal thickness (in millimeters), H is the hydration (milligrams water per milligrams dry tissue weight), and a and b are constants depending on the type of species. In human cornea, for example, a = 7.0 and b = 0.64, whereas in New Zealand small rabbits (<4 kg) a = 10.0 and b = 0.42 have been suggested. 32  
Results
Figure 1 depicts typical OCT images of the rabbit cornea before induction of dehydration (Fig. 1a) , at the peak of dehydration (Fig. 1b) , and at the peak of rehydration (Fig. 1c) after topical application of the hypertonic agent. 
The corresponding OCT signals monitored by oscilloscope are also provided below the images. The oscilloscope signal in the area between the peaks of reflection from the epithelium and endothelium (marked with short arrows in Fig. 1 ) represents the backscatter of the stroma, and the larger value of this signal in the dehydrated state corresponds to the brighter bands that appear in the OCT images of the anterior section of the cornea (compare Fig. 1b with 1a and 1c ). The considerable decrease in the time delay between the two reflection peaks from the corneal boundaries after application of the hypertonic agent exhibits a proportional decrease of thickness in the corresponding oscilloscope plot shown below this image (Fig. 1b) . Also, the increased scattering can easily be detected and quantified in the same plot. The maximum amplitude of the backscatter in Figure 1b is approximately 0.4 V higher than the amplitude in Figure 1c at the same depth. That means there is a change of approximately 18 dB in logarithmic scale (which translates to a dramatic increase, up to eight times in linear scale) between the amplitude of the signal corresponding to a totally dehydrated state and that of the overhydrated cornea. There is also a considerable change in the profile of the backscatter in-depth distribution that is dependent on the hydration state of the cornea. The maximum backscatter was found to be located within a 150-μm depth below the epithelial layer of the cornea. We introduced two slopes to characterize this profile quantitatively (front and back as depicted by straight lines in Fig. 1 ). These slopes are steeper in the dehydrated state (Fig. 1b) . In Figure 1c , the front slope changes from a positive value to negative value due to overhydration. 
Figure 2 exhibits the dehydration and rehydration cycles of rabbit cornea treated with different concentrations (50% and 100%) of the hypertonic agent. A plot of corneal thickness versus time provided a quantitative means to assess the dynamic CCT change during drug-induced dehydration of the cornea. At the same time, the profile of light scattering within the cornea could be visualized in the OCT images. Depending on the state of the hydration of the cornea and the dehydrating procedures used in the experiments, we observed a minimum and maximum CCT of approximately 200 and 700 μm, respectively (i.e., a CCT change of up to 500 μm caused by the dehydrating effects of the hypertonic agent and the subsequent overrehydration of the treated cornea). The two graphs in Figure 2 demonstrate the influence of drug concentration on the rate and extent of dehydration and rehydration of the cornea after topical application of the hypertonic agent. Corneas treated with the 100% hypertonic agent exhibited an initial rehydration rate of 19 μm/min compared with 10 μm/min for the 50% hypertonic agent. As shown in Figure 2 , the cornea overshoots its initial thickness during the rehydration process. The cornea treated with 50% hypertonic agent showed an overshoot of 40% above its original thickness, whereas the cornea treated with 100% hypertonic agent concentration exhibited an overshoot of up to 90% above its initial thickness. For both hypertonic agent concentrations, the temporal responses of the cornea to the first dehydrating stress followed by the second dehydrating stress that was applied after recovery of the cornea from the first stress are depicted in Figure 2 . These results show a slower rate of corneal recovery after the second dehydrating stress than that of the cornea’s recovery rate after the application of the first dehydrating stress. 
Table 1 presents the data for the CCT recovery in four rabbits after dehydration by the hypertonic agent and shows the range of parameters and accuracy of the fit provided by the proposed exponential model (i.e., equation 1a ). These data include the initial corneal thickness, the extent of corneal thinning induced by the application of hypertonic agent (100% concentration), the experimentally measured overshoot after the stress test, and the statistical parameters of exponential fit. From these data we calculated the average time constant for the recovery process as 23 minutes, with an average initial recovery rate of 12.4 μm/min based on equation 2a
The observed rehydration overshoot was reversible over time. Seven days after the dehydration stress studies on the rabbit corneas, we grossly examined the treated corneas and measured their thickness with our OCT system. The corneas had returned to their original thicknesses and showed no sign of edema. 
Figure 3 depicts the decrease in corneal thickness during prolonged exposure of the cornea to quiescent air at room temperature. Use of equation 2b showed the estimated initial rate of the decrease in corneal thickness to be 2.7 μm/min. After the eye was closed, it took approximately 1 hour for the cornea to recover. Thus, the average rate of the recovery process is estimated to be approximately 1.5 μm/min. 
Table 2 contains the range, weighted means, and standard deviations of the time constant for the dehydration cycles caused by the 100% hypertonic agent and the free surface evaporation at room temperature. The average induced thinning was approximately 105 μm in the hypertonic agent–treated group. Use of equation 2b showed an estimated average rate of 74 μm/min for dehydration-induced changes in CCT. 
Figure 4 demonstrates the ability of OCT to provide quantitative details about stress-induced changes in the scattering properties of the cornea. In this study, changes in scattering properties of the stromal tissue during the dehydration and rehydration cycles were monitored. The correlation between changes in corneal thickness and the maximum scattering intensity as a function of time after topical application of the hypertonic agent (100% concentration) is shown in Figure 4 . Our analyses show that intensity of the backscattered light varied up to 20 times (25 ±1 dB) between the peaks of the de- and rehydration phases. The uncertainty in the quantification of dehydration- and rehydration-induced changes in the degree of corneal backscatter is ±1 dB. 
Discussion
In the past, many studies have been conducted to investigate CHC dynamics 33 34 35 36 and the relationship between corneal transparency and the extent of corneal hydration. In most of the studies, corneal swelling is induced by hypoxia, and the rate of corneal de-swelling is quantified by measuring corneal thickness as a function of time. There has been considerable interest in developing techniques to quantify the amount and the spatial distribution of water in the cornea. Most of the proposed techniques for direct measurement of corneal hydration require utilization of invasive procedures. 37 38 In recent years, a noninvasive means such as pachymetry 39 has been developed for clinical use to measure CCT and to infer information about the state of corneal hydration with an indirect approach. However, these measurements often require direct contact of a probe with the cornea and generally do not provide any details about the spatial distribution of water within the cornea. In the present study, corneal thinning was induced, and then the rate of recovery of swelling was investigated while the changes in corneal thickness and backscatter were monitored. 
In the past, clinical measurements of light-scattering within the cornea have been suggested for the study of corneal hydration and its dynamics. 40 The appearance of bright bands in the OCT image of the anterior part of the cornea in the dehydrated state (Fig. 1b) , indicates the qualitative change of the scattering profile across this layer as a result of water movement within the cornea. The local depletion of water within the cornea is known to induce geometric changes between collagen fibrils and to distort the fine hexagonal arrangement between these fibrils that normally ensures the transparency of the cornea. 23 The relatively high amplitude of the backscatter in the anterior one third of the cornea suggests lower water content in this region compared with the other two thirds of the cornea. The change in the hydration state of the cornea also alters the slopes of the backscatter in-depth distribution in the stromal part, as shown in Figure 1 , in a reproducible manner. Furthermore, the spatial distribution of the backscatter from the stromal layer correlates inversely with the known water distribution in the cornea, as described by Turss et al. 41  
The descending slopes in Figure 2 show the rate of corneal dehydration caused by topical application of two different concentrations of Ophthalgan, the dehydrating agent. It also demonstrates the strong response of the cornea to the dehydration stress by the second ascending slope (rehydration one). This is dependent on the severity of this stress, as evidenced by the speed of recovery of corneal thickness. Comparison of the rate of corneal dehydration after topical application of the hypertonic agent to the rate caused by evaporation on prolonged exposure of the cornea to quiescent air at room temperature (Fig. 3) reveals a significant difference in the speed of the corneal reaction to different dehydrating methods. In the case of drug-induced dehydration, rapid dehydration is followed by a quick rehydration response and an overshoot in corneal hydration, as shown in the thickening of the cornea beyond its initial unperturbed thickness. The magnitude of the overshoot in corneal thickness depends on the amount of dehydrating stress (Fig. 2) for 100% versus 50% hypertonic agent concentrations. This can also be estimated by observing the disparity in the time constants (Table 2) given for the different dehydrating means. It is interesting to note that an overshoot in human cornea during recovery to its original thickness after a hypoxia stress test was also reported previously. 34 35 42 However, in those reports, the overshoot was captured in the dehydrating process from an edematous (overhydrated) condition caused by induced hypoxia toward the initial unperturbed corneal thickness. In our study, the overshoot occurred in the rehydrating process from a dehydrated state and went beyond the initial corneal thickness. 
Figure 4 shows that the intensity of the backscatter correlates inversely with the corneal thickness and hence qualitatively with the hydration of the cornea. From the data presented in Figure 4 , a linear relationship between the corneal thickness L (in millimeters) and the magnitude of backscatter (S, in decibels) from the cornea in vivo was found by fitting the experimental results depicted in Figure 4 with a linear fit (R = −0.83, P < 0.0001):  
\[S(\mathrm{dB})\ {=}\ (60.3\ {\pm}\ 4.6)\ {-}\ (71\ {\pm}\ 10)L(\mathrm{mm})\]
 
Subsequently, by considering equation 3 reported by Hedbys and Mishima 32 and the observed relationship stated in equation 4 between the magnitude of backscatter S (in decibels) and corneal thickness, one can derive a direct relationship between the magnitude of the backscatter and corneal hydration (H), which can be stated as the following equation:  
\[S(\mathrm{dB})\ {=}\ (57.3\ {\pm}\ 5.0)\ {-}\ (7.1\ {\pm}\ 1.0)\ H\]
This equation can be used for in vivo assessment of the corneal hydration, thus allowing for noninvasive quantitative monitoring of corneal hydration using OCT. From equation 5 , one can see that the degree of backscatter light as detected by OCT decreases as cornea hydration increases. To explain this observation, one should consider the lattice structure of the cornea and the diffractive nature of corneal transparency, as originally proposed by Maurice. 23 In the infrared region, during the process of corneal overhydration, the magnitude of the OCT signal is initially influenced by the lattice structure of the cornea. However, as the extent of corneal swelling is increased and the lattice structure of the cornea is disturbed, the wide-angle scattering makes a larger contribution to the detected OCT. Thus, in our view, the observed initial decrease in the OCT signal is primarily due to a swelling-induced decrease in the number of lattice elements within a small volume of tissue that is probed by OCT. 
The observed overshoot can be attributed to the fact that the hypertonic agent used in our experiments can penetrate readily across the cornea, reach the endothelial layer, and affect the pumping action of the endothelial cells. 43 Thus, the direct effects of this hypertonic agent on the endothelial pump function can compromise the function of endothelial cells temporarily. As more water is pulled into the cornea, the observed reversible overhydration and thickening of the cornea after topical application of the hypertonic agent is seen. Although transient reaction of the epithelial layer to glycerin is partly responsible for the observed response of the cornea during the dehydration stress, the response of the cornea to the dehydrating agent is predominantly due to the interaction of glycerin with the stromal tissue and endothelial function. Several cryobiological studies have already investigated the effects of different concentrations of glycerin (the main chemical ingredient of this hypertonic agent) on the cornea and reported findings that support our proposition that the paralyzing effect of the glycerin on endothelial pump function 44 45 46 47 48 could be responsible for the induction of the overshoot in corneal thickness after the application of hypertonic agent. As mentioned earlier, after the in vivo experiments, the corneas recovered to their original thicknesses and were free of edema 1 week after each experiment. This suggests the reversibility of the observed overshoot and the recovery of the endothelial pump function over time. The ability of this hypertonic agent to induce reversible changes in corneal morphology and optical properties as quantified by OCT monitoring in vivo is also consistent with the findings of other studies that investigated the effects of different concentrations of glycerin on the endothelial layer of the cornea in vitro. 45 46 Collectively, the current in vivo investigation as well as the prior in vitro studies suggests that a stress test using the optimum concentration of dehydrating agent combined with high-resolution OCT measurements can be developed for the quantitative assessment of corneal function and viability. 
In addition, it’s worth noting that multiwavelength differential OCT measurements can expand the utility of the proposed monitoring technique by mapping the distribution of scatterers and absorbers within the cornea to determine water distribution within an unperturbed cornea. 
Conclusion
OCT monitoring could provide a safe, noninvasive, and quantitative method for studying the dynamics of corneal response after topical application of a dehydrating agent. Investigation of corneal backscatter distribution using this technique appears to offer promising prospects for quantitative measurements of the hydration gradient inside the cornea. Thus, quantitative OCT imaging could open a new perspective for noninvasive assessment of corneal morphology, function, and viability. 
 
Figure 1.
 
Images of the unperturbed (a), dehydrated (b), and overhydrated (c) states of rabbit cornea in vivo. The corresponding oscilloscope plot is presented below each image. Temporal averaging over 128 scans and spatial averaging over an area of 100 × 100 μm reduced the noise in the oscilloscope graphs. Short arrows: indicate the reflection peaks from the epithelium and endothelium. Diamond arrows: show front and back slopes. The single oval arrow in each graph shows an artifact caused by internal optical reflections in the OCT system.
Figure 1.
 
Images of the unperturbed (a), dehydrated (b), and overhydrated (c) states of rabbit cornea in vivo. The corresponding oscilloscope plot is presented below each image. Temporal averaging over 128 scans and spatial averaging over an area of 100 × 100 μm reduced the noise in the oscilloscope graphs. Short arrows: indicate the reflection peaks from the epithelium and endothelium. Diamond arrows: show front and back slopes. The single oval arrow in each graph shows an artifact caused by internal optical reflections in the OCT system.
Figure 2.
 
The changes in corneal thicknesses during the de- and rehydration phase of corneal response to topical application of hypertonic agent: 100% (•), 50% (○). The hydration level is indicated on the right y-axis.
Figure 2.
 
The changes in corneal thicknesses during the de- and rehydration phase of corneal response to topical application of hypertonic agent: 100% (•), 50% (○). The hydration level is indicated on the right y-axis.
Table 1.
 
Rehydration (Closed Eye) Subsequent to Dehydration Caused by Application of Hypertonic Agent
Table 1.
 
Rehydration (Closed Eye) Subsequent to Dehydration Caused by Application of Hypertonic Agent
Animal Initial Corneal Thickness* (μm) Induced Thinning* (μm) Experimental Overshoot* (μm) Calculated Maximum Overshoot (μm) Calculated Time Constant (T hyd) (min) Calculated Maximum Recovery Rate (dL/dt), μm/min
1 310 −124 49 98 ± 32 20.4 ± 7.5 8.0 ± 3.0
2 309 −65 37 37 ± 19 14.0 ± 1.4 7.4 ± 0.7
3 319 −75 228 322 ± 38 28.1 ± 3.2 14.3 ± 2.0
4 355 −156 316 414 ± 47 28.8 ± 3.5 19.8 ± 1.6
Figure 3.
 
The changes in corneal thicknesses during the de- and rehydration process after corneal exposure to quiescent air as the dehydrating process. The hydration level is indicated on the right y-axis.
Figure 3.
 
The changes in corneal thicknesses during the de- and rehydration process after corneal exposure to quiescent air as the dehydrating process. The hydration level is indicated on the right y-axis.
Table 2.
 
The Estimated Time Constants and the Magnitude of the Induced Thinning for Corneal Response during Dehydration Processes for both Hypertonic Agent and Free Surface Evaporation
Table 2.
 
The Estimated Time Constants and the Magnitude of the Induced Thinning for Corneal Response during Dehydration Processes for both Hypertonic Agent and Free Surface Evaporation
Process Animals (n) Mean Value of Time Constant (T deh), Standard Deviation and Range* (min) Mean Value of Induced Thinning, Standard Deviation and Range* (μm)
Hypertonic agent-induced dehydration 4 1.41 ± 0.07 (0.9–1.8) 105 ± 39 (65–156)
Natural dehydration (evaporation in air) 4 12.5 ± 0.8 (8.7–19.6) 60.6 ± 3.4 (37–89)
Figure 4.
 
Changes of thickness (•) and backscatter intensity (○) of the cornea as a function of time during dehydration (by 100% Ophthalgan; Wyeth) and rehydration cycles. Solid and dashed lines: interpolation of experimental data for thickness and corneal backscatter, respectively.
Figure 4.
 
Changes of thickness (•) and backscatter intensity (○) of the cornea as a function of time during dehydration (by 100% Ophthalgan; Wyeth) and rehydration cycles. Solid and dashed lines: interpolation of experimental data for thickness and corneal backscatter, respectively.
The authors thank Kirill Larin and Brent Bell for technical assistance. 
Ruberti JW, Klyce SD. NaCl osmotic perturbation can modulate hydration control in rabbit cornea. Exp Eye Res. 2003;76:349–359. [CrossRef] [PubMed]
Pallikaris IG, Ginis HS, Kounis GA, Anglos D, Papazoglou TG, Naoumidis LP. Corneal hydration monitored by laser-induced breakdown spectroscopy. J Refract Surg. 1998;14:655–660. [PubMed]
Bauer NJ, Wicksted JP, Jongsma FH, March WF, Hendrikse F, Motamedi M. Noninvasive assessment of the hydration gradient across the cornea using confocal Raman spectroscopy. Invest Ophthalmol Vis Sci. 1998;39:831–835. [PubMed]
Kim WS, Jo JM. Corneal hydration affects ablation during laser in situ keratomileusis surgery. Cornea. 2001;20:394–397. [CrossRef] [PubMed]
Kohlhaas M, Draeger J, Bohm A. Effect of corneal hydration and dehydration of donor tissue on the precision of the thickness in lamellar refractive keratoplasty. Ophthalmologe. 1995;92:410–413. [PubMed]
Aboalchamat B, Engelmann K, Bohnke M, Eggli P, Bednarz J. Morphological and functional analysis of immortalized human corneal endothelial cells after transplantation. Exp Eye Res. 1999;69:547–553. [CrossRef] [PubMed]
Doughty MJ, Zaman ML. Human corneal thickness and its impact on intraocular pressure measures: a review and meta-analysis approach. Surv Ophthalmol. 2000;44:367–408. [CrossRef] [PubMed]
Simon G, Small RH, Ren Q, Parel JM. Effect of corneal hydration on Goldmann applanation tonometry and corneal topography. Refract Corneal Surg. 1993;9:110–117. [PubMed]
Lleo A, Marcos A, Calatayud M, Alonso L, Rahhal SM, Sanchis-Gimeno JA. The relationship between central corneal thickness and Goldmann applanation tonometry. Clin Exp Optom. 2003;86:104–108. [CrossRef] [PubMed]
Egan CA, Hodge DO, McLaren JW, Bourne WM. Effect of dorzolamide on corneal endothelial function in normal human eyes. Invest Ophthalmol Vis Sci. 1998;39:23–29. [PubMed]
Wirtitsch MG, Findl O, Kiss B, Petternel V, Heinzl H, Drexler W. Short-term effect of dorzolamide hydrochloride on central corneal thickness in humans with cornea guttata. Arch Ophthalmol. 2003;121:621–625. [CrossRef] [PubMed]
Koozekanani D, Boyer K, Roberts C. Retinal thickness measurements from optical coherence tomography using a Markov boundary model. IEEE Trans Med Imaging. 2001;20:900–916. [CrossRef] [PubMed]
Neubauer AS, Priglinger SG, Thiel MJ, May CA, Welge-Lussen UC. Sterile structural imaging of donor cornea by optical coherence tomography. Cornea. 2002;21:490–494. [CrossRef] [PubMed]
Hirano K, Ito Y, Suzuki T, Kojima T, Kachi S, Miyake Y. Optical coherence tomography for the noninvasive evaluation of the cornea. Cornea. 2001;20:281–289. [CrossRef] [PubMed]
Wirbelauer C, Winkler J, Bastian GO, Haberle H, Pham DT. Histopathological correlation of corneal diseases with optical coherence tomography. Graefes Arch Clin Exp Ophthalmol. 2002;240:727–734. [CrossRef] [PubMed]
Welzel J. Optical coherence tomography in dermatology: a review. Skin Res Technol Rev. 2001;7:1–9. [CrossRef]
Chan KP, Devaraj B, Yamada M, Inaba H. Coherent detection techniques in optical imaging of tissues. Phys Med Biol. 1997;42:855–867. [CrossRef] [PubMed]
Roper SN, Moores MD, Gelikonov GV, et al. In vivo detection of experimentally induced cortical dysgenesis in the adult rat neocortex using optical coherence tomography. J Neurosci Methods. 1998;80:91–98. [CrossRef] [PubMed]
Kholodnykh A, Hosseini K, Petrova I, Essenaliev R, Motamedi M. In vivo OCT assessment of rabbit corneal hydration and dehydration. Proc SPIE. 2003;4956:295–298.
Kholodnykh AI, Petrova IY, Larin KV, Motamedi M, Esenaliev RO. Optimization of low coherence interferometry for quantitative analysis of tissue optical properties. Proc SPIE. 2002;4624:36–46.
Kholodnykh AI, Petrova IY, Larin KV, Motamedi M, Esenaliev RO. Precision of measurement of tissue optical properties with optical coherence tomography. Appl Opt. 2003;42:3027–3037. [CrossRef] [PubMed]
Aubert H, Matthiesen L. Graefe-Saemisch Handbuch der gesammten Augenheilkunde. 1876; 1st ed. 409. Engelmann Leipzig, Germany.
Maurice DM. The structure and transparency of the cornea. J Physiol. 1957;136:263–286. [CrossRef] [PubMed]
Goodman JW. Some fundamental properties of speckles. J Opt Soc Am. 1976;66:1145–1150. [CrossRef]
Schmitt JM. Speckle in optical coherence tomography. J Biomed Optics. 1999;4:95–105. [CrossRef]
Luxenberg MN, Green K. Reduction of corneal edema with topical hypertonic agents. Am J Ophthalmol. 1971;71:847–853. [CrossRef] [PubMed]
Polse KA, Brand R, Mandell R, Vastine D, Demartini D, Flom R. Age differences in corneal hydration control. Invest Ophthalmol Vis Sci. 1989;30:392–399. [PubMed]
Mandell RB, Polse KA, Brand RJ, Vastine D, Demartini D, Flom R. Corneal hydration control in Fuchs’ dystrophy. Invest Ophthalmol Vis Sci. 1989;30:845–852. [PubMed]
Herse PR. Corneal hydration control in normal and alloxan-induced diabetic rabbits. Invest Ophthalmol Vis Sci. 1990;31:2205–2213. [PubMed]
Weston BC, Bourne WM, Polse KA, Hodge DO. Corneal hydration control in diabetes mellitus. Invest Ophthalmol Vis Sci. 1995;36:586–595. [PubMed]
Matsuura T, Gorti S, Tanaka T, Hara Y, Saishin M. Determination of corneal gel dynamics. Eur Biophys J. 1999;28:357–368. [CrossRef] [PubMed]
Hedbys BO, Mishima S. The thickness-hydration relationship of the cornea. Exp Eys Res. 1966;5:221–228. [CrossRef]
Maurice DM. The cornea and sclera. Davson H eds. 3rd ed. The Eye. 1984;1b:1–158. Academic Press New York.
O’Neal MR, Polse KA. In vivo assessment of mechanisms controlling corneal hydration. Invest Ophthalmol Vis Sci. 1985;26:849–856. [PubMed]
Bourassa S, Benjamin WJ, Boltz RL. Effect of humidity on the deswelling function of the human cornea. Curr Eye Res. 1991;10:493–500. [CrossRef] [PubMed]
Siu AW, Herse PR. The effect of age on the edema response of the central and midperipheral cornea. Acta Ophthalmol. 1993;71:57–61.
Castoro JA, Bettelheim AA, Bettelheim FA. Water gradients across bovine cornea. Invest Ophthalmol Vis Sci. 1988;29:963–968. [PubMed]
Monti D, Chetoni P, Burgalassi S, Najarro M, Saettone MF. Increased corneal hydration induced by potential ocular penetration enhancers: assessment by differential scanning calorimetry (DSC) and by desiccation. Int J Pharm. 2002;31(232)139–147.
La Rosa FA, Gross RL, Orengo-Nania S. Central corneal thickness of Caucasians and African Americans in glaucomatous and nonglaucomatous populations. Arch Ophthalmol. 2001;119:23–27. [PubMed]
Olsen TH. Light scattering from the human cornea. Invest Ophthalmol Vis Sci. 1982;23:81–86. [PubMed]
Turss R, Friend J, Reim M, Dohlman CH. Glucose concentration and hydration of the corneal stroma. Ophthalmic Res. 1971;2:253–260. [CrossRef]
Odenthal MTP, Niewuendaal CP, Venema HW, Oosting J, Kok JHC, Kijlstra A. In vivo human corneal hydration control dynamics: a new model. Invest Ophthalmol Vis Sci. 1999;40:312–319. [PubMed]
Grass GM, Robinson JR. Mechanisms of corneal drug penetration. I: In vivo and in vitro kinetics. J Pharm Sci. Jan 1988;77:3–14. [CrossRef] [PubMed]
Bourne WM, Shearer DR, Nelson LR. Human corneal endothelial tolerance to glycerol, dimethylsulfoxide, 1,2-propanediol, and 2,3-butanediol. Cryobiology. 1994;31:1–9. [CrossRef] [PubMed]
Rich SJ, Armitage WJ. Corneal tolerance of vitrifiable concentrations of glycerol. Cryobiology. 1992;29:153–164. [CrossRef] [PubMed]
Rich SJ, Armitage WJ. The potential of an equimolar combination of propane-1,2-diol and glycerol as a vitrification solution for corneas. Cryobiology. 1991;28:314–326. [CrossRef] [PubMed]
Brunette I, Nelson LR, Bourne WM. Tolerance of human corneal endothelium to glycerol. Cryobiology. 1989;26:513–523. [CrossRef] [PubMed]
Madden PW. The assessment of endothelial integrity by scanning electron microscopy and fluorescein diacetate staining following treatment with cryoprotective additives. Curr Eye Res. 1989;8:17–36. [CrossRef] [PubMed]
Taylor JR. An Introduction to Error Analysis. 1997; 2nd ed. 173–179. University Science Books Sausalito, CA.
Figure 1.
 
Images of the unperturbed (a), dehydrated (b), and overhydrated (c) states of rabbit cornea in vivo. The corresponding oscilloscope plot is presented below each image. Temporal averaging over 128 scans and spatial averaging over an area of 100 × 100 μm reduced the noise in the oscilloscope graphs. Short arrows: indicate the reflection peaks from the epithelium and endothelium. Diamond arrows: show front and back slopes. The single oval arrow in each graph shows an artifact caused by internal optical reflections in the OCT system.
Figure 1.
 
Images of the unperturbed (a), dehydrated (b), and overhydrated (c) states of rabbit cornea in vivo. The corresponding oscilloscope plot is presented below each image. Temporal averaging over 128 scans and spatial averaging over an area of 100 × 100 μm reduced the noise in the oscilloscope graphs. Short arrows: indicate the reflection peaks from the epithelium and endothelium. Diamond arrows: show front and back slopes. The single oval arrow in each graph shows an artifact caused by internal optical reflections in the OCT system.
Figure 2.
 
The changes in corneal thicknesses during the de- and rehydration phase of corneal response to topical application of hypertonic agent: 100% (•), 50% (○). The hydration level is indicated on the right y-axis.
Figure 2.
 
The changes in corneal thicknesses during the de- and rehydration phase of corneal response to topical application of hypertonic agent: 100% (•), 50% (○). The hydration level is indicated on the right y-axis.
Figure 3.
 
The changes in corneal thicknesses during the de- and rehydration process after corneal exposure to quiescent air as the dehydrating process. The hydration level is indicated on the right y-axis.
Figure 3.
 
The changes in corneal thicknesses during the de- and rehydration process after corneal exposure to quiescent air as the dehydrating process. The hydration level is indicated on the right y-axis.
Figure 4.
 
Changes of thickness (•) and backscatter intensity (○) of the cornea as a function of time during dehydration (by 100% Ophthalgan; Wyeth) and rehydration cycles. Solid and dashed lines: interpolation of experimental data for thickness and corneal backscatter, respectively.
Figure 4.
 
Changes of thickness (•) and backscatter intensity (○) of the cornea as a function of time during dehydration (by 100% Ophthalgan; Wyeth) and rehydration cycles. Solid and dashed lines: interpolation of experimental data for thickness and corneal backscatter, respectively.
Table 1.
 
Rehydration (Closed Eye) Subsequent to Dehydration Caused by Application of Hypertonic Agent
Table 1.
 
Rehydration (Closed Eye) Subsequent to Dehydration Caused by Application of Hypertonic Agent
Animal Initial Corneal Thickness* (μm) Induced Thinning* (μm) Experimental Overshoot* (μm) Calculated Maximum Overshoot (μm) Calculated Time Constant (T hyd) (min) Calculated Maximum Recovery Rate (dL/dt), μm/min
1 310 −124 49 98 ± 32 20.4 ± 7.5 8.0 ± 3.0
2 309 −65 37 37 ± 19 14.0 ± 1.4 7.4 ± 0.7
3 319 −75 228 322 ± 38 28.1 ± 3.2 14.3 ± 2.0
4 355 −156 316 414 ± 47 28.8 ± 3.5 19.8 ± 1.6
Table 2.
 
The Estimated Time Constants and the Magnitude of the Induced Thinning for Corneal Response during Dehydration Processes for both Hypertonic Agent and Free Surface Evaporation
Table 2.
 
The Estimated Time Constants and the Magnitude of the Induced Thinning for Corneal Response during Dehydration Processes for both Hypertonic Agent and Free Surface Evaporation
Process Animals (n) Mean Value of Time Constant (T deh), Standard Deviation and Range* (min) Mean Value of Induced Thinning, Standard Deviation and Range* (μm)
Hypertonic agent-induced dehydration 4 1.41 ± 0.07 (0.9–1.8) 105 ± 39 (65–156)
Natural dehydration (evaporation in air) 4 12.5 ± 0.8 (8.7–19.6) 60.6 ± 3.4 (37–89)
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×