Optical Coherence Tomography (OCT) is a noninvasive imaging modality that has significantly contributed to the quantitative assessment of ocular diseases. Another tool available to ophthalmic clinicians is in-vivo confocal microscopy, which allows anatomical structures to be observed live at the cellular level. Incorporating both of these modalities for imaging the cornea allows us to take structural measurements to characterize disease-related changes in corneal anatomy. Notable diseases that directly impact or correlate with corneal structures include glaucoma and diabetic neuropathy. Given glaucoma's impact as the second leading cause of blindness in the world, great efforts have been made in researching and understanding the disease. Correlations have been found between the central corneal thickness (CCT) and the risk of developing visual field loss in patients diagnosed with glaucoma. It should come as no surprise that measuring CCT among glaucoma suspects has also now become a clinical standard of practice. Diabetes is a group of metabolic diseases where the body experiences high blood sugar levels over prolonged periods of time. It is a prominent disease that affects millions of Americans each day. While not necessarily an ocular disease in its own right, it has been shown that diabetes can still affect the corneal structures. Diabetics have decreased corneal sensitivity and a significant link has been established between neuropathic severity in diabetic patients and corneal nerve fiber density. Given the availability of these imaging tools and the significant impact these prominent diseases have on society a growing focus has developed on relating corneal structure measurements and ophthalmic diseases. However, manually acquiring structural measures of the cornea can be a labor intensive and daunting task. Hence, experts have sought to develop automatic alternatives. The goals of our work includes the ability to automatically segment the corneal structures from anterior segment-optical coherence tomography (AS-OCT) and in-vivo confocal microscopy (IVCM) to provide useful structural information from the cornea. The major contributions of this work include 1) utilizing the information of AS-OCT imagery to segment the cornea layers simultaneously in 3D, 2) increasing the region-of-interest of IVCM imagery using a feature-based registration approach to develop a panorama from the images, 3) incorporating machine-learning techniques to segment the corneal nerves in the IVCM imagery, and 4) extracting structural measurements from the segmentation results to determine correlations between the structural measurements known to differ from the corneal structures in various subject groups.

Corneal ulceration with prolonged healing following Nd:YAG laser cyclophotocoagulation in dogs is a frequent complication. It is hypothesized that these corneal ulcerations may be a form of neurotrophic keratitis due to laser-induced damage to corneal innervation. Fifteen clinically normal dogs had the neodymium:yttrium aluminum garnet(Nd:YAG) laser cyclophotocoagulation performed on the left eye. Each treated eye received 100 Joules of laser energy. Corneal touch threshold (CTT) and Schirmer I tear tests (STT) were performed before the surgery and on days 1,3,5,7,9,11, and 13 post-laser treatment. Applanation tonometry was performed before surgery and twice daily for 14 days post-laser treatment. Eyes were enucleated after 14 days and corneal nerves were stained using a gold chloride technique. Major nerve bundles entering the cornea were quantitated by quadrant, using camera lucida reproductions. Nerve bundle diameters were measured using NIH image computer software on computer-scanned images. Statistical methods included repeated values for analysis of variance for CTT, STT and IOP, and a paired t-test for nerve diameters and bundles. All laser treated eyes had significantly higher CTTs (P<0.05) compared to control eyes for all measurements. Six out of fifteen dogs had evidence of ulcerative keratitis. Intraocular pressure was significantly lower in laser treated eyes compared to control eyes in the a.m. on days 2-9,and 14, and in the p.m. on days 2-11 using a Bonferroni-corrected alpha level (P<0.0039). A significant decrease of one nerve bundle per corneal quadrant was found between the laser treated and control eyes. There was no significant difference in STT or nerve bundle diameters between laser treated and control eyes. Nd:YAG laser cyclophotocoagulation effectively reduces IOP while increasing CTT. The procedure also causes a significant decrease in the number of major nerve bundles entering the cornea, but has no effect on the diameter of those bundles. These findings support the hypothesis that nerve damage and corneal hypoesthesia are etiologic factors in ulcerative keratitis following Nd:YAG laser cyclophotocoagulation. / Master of Science

Notre projet vise à élucider la relation complexe entre la morphologie des nerfs cornéens et la sensibilité cornéenne, afin d'améliorer la compréhension et le diagnostic des pathologies oculaires. En utilisant deux types d'esthésiomètres : l'esthésiomètre sans contact (NCCA) et le Cochet-Bonnet (CBA) pour mesurer la sensibilité, et en analysant les images de microscopie confocale (IVCM) via le logiciel CCMetrics, nous avons étudié 23 individus, y compris ceux souffrant de diabète et de kératite neurotrophique. Des corrélations négatives significatives entre certains attributs neuronaux et la sensibilité cornéenne ont été identifiées. L'utilisation d'algorithmes d'apprentissage automatique, tels que K-Plus Proches Voisins (KNN), les Réseaux de Neurones (MLP), la Régression à Vecteurs de Support (SVR) et les arbres de décision, a révélé des relations non linéaires complexes. Notre étude encourage l'utilisation de l’apprentissage automatique pour détecter ces relations complexes dans le domaine médical en général et en ophtalmologie en particulier. / Our project aims to clarify the complex relationship between the morphology of corneal nerves and corneal sensitivity, to improve understanding and diagnosis of eye pathologies. We used two types of esthesiometers: a non-contact esthesiometer (NCCA) and Cochet-Bonnet (CBA) for sensitivity measurement and analyzed confocal microscopy (IVCM) images using the software CCMetrics. We studied 23 individuals, including those with diabetes and neurotrophic keratitis. Significant negative correlations between certain neuronal attributes and corneal sensitivity were identified. The use of machine learning algorithms, such as K-Nearest Neighbors (KNN), Neural Networks (MLP), Support Vector Regression (SVR), and decision trees, revealed complex non-linear relationships. Our study advocates using machine learning to detect these complex relationships in the medical field, especially in ophthalmology.

Introduction: Relation of diabetes mellitus (DM) to the diabetic keratopathy and various stages of corneal nerve fiber damage has been well accepted. A possible association between changes in the cornea of diabetic patients and diabetic retinopathy (DR), DM duration, and age at the time of DM diagnosis were evaluated. Neuropathies are among the most common long-term complications of diabetes mellitus. Good glycemic control is essential in prevention of this complication. DM patients with similar mean glucose levels or glycated hemoglobin (HbA1c) levels often exhibit differences in evaluation of diabetic complications. One reason for these differences may be the differences in glucose variability. DM patients with similar mean glucose levels or HbA1c levels often exhibit differences in glucose variability Hypothesis: Diabetes mellitus damages the subbasal nerve fibers of the corneal and affects the density of epithelial, endothelial and stromal cells. Corneal changes in patients with DM are dependent on the degree of diabetic retinopathy (DR), age at diagnosis, duration of DM, and compensation parameters. Purpose: To compare changes in cell density in individual layers of cornea and status of subbasal nerve fibers in patients with type 1 DM (DM 1) and in healthy subjects. To evaluate the dependence...