Ocular pathology is one of the main health problems worldwide. The number of people with retinopathy symptoms has increased considerably in recent years. Early adequate treatment has demonstrated to be effective to avoid the loss of the vision. The analysis of fundus images is a non intrusive option for periodical retinal screening. Different models designed for the analysis of retinal images are based on supervised methods, which require of hand labelled images and processing time as part of the training stage. On the other hand most of the methods have been designed under the basis of specific characteristics of the retinal images (e.g. field of view, resolution). This compromises its performance to a reduce group of retinal image with similar features. For these reasons an unsupervised model for the analysis of retinal image is required, a model that can work without human supervision or interaction. And that is able to perform on retinal images with different characteristics. In this research, we have worked on the development of this type of model. The system locates the eye structures (e.g. optic disc and blood vessels) as first step. Later, these structures are masked out from the retinal image in order to create a clear field to perform the lesion detection. We have selected the Graph Cut technique as a base to design the retinal structures segmentation methods. This selection allows incorporating prior knowledge to constraint the searching for the optimal segmentation. Different link weight assignments were formulated in order to attend the specific needs of the retinal structures (e.g. shape). This research project has put to work together the fields of image processing and ophthalmology to create a novel system that contribute significantly to the state of the art in medical image analysis. This new knowledge provides a new alternative to address the analysis of medical images and opens a new panorama for researchers exploring this research area.

The general aim of the thesis was to deduce whether selected naturally occurring flavonoids (genistein, epicatechin gallate (EC), epigallocatechin gallate (EGCG), baicalin) attenuate various secondary insults that may cause death of ganglion cells in primary open angle glaucoma (POAG). An ischemic insult to the rat retina significantly causes the inner retina to degenerate indexed by changes of various antigens, proteins and mRNAs located to amacrine and ganglion cells. These changes are blunted in animals treated with genistein as has been shown for ECGC. Studies conducted on cells (RGC-5 cells) in culture showed that hydrogen peroxide, L-buthionine sulfoximine (BSO)/glutamate and serum deprivation (mimicking oxidative stress), rotenone, sodium azide (affecting mitochondria function in specific ways) and light (where the mitochondria are generally affected) all generated reactive oxygen species and caused death of RGC-5 cells. EGCG was able to attenuate cell death caused by hydrogen peroxide, sodium azide and rotenone. Only EC was able to attenuate BSO/glutamate-induced cell death, in addition to cell death caused by hydrogen peroxide and rotenone. Genistein had no positive effect on cell death in experiments carried out on RGC-5 cells. Exposure of RGC-5 cells to flavonoids showed that EC and EGCG increased the mRNA expression of endogenous antioxidants such as HO-l (heme oxygenase 1) and Nrf-2 (nuclear erythroid factor-z-related factor 2). Light insult, rotenone and sodium azide activate the p38 (protein kinase 38) pathway, while only light and rotenone activate the JNK (c-Jun amino-terminal kinase) pathway. Serum deprivation affects mitochondrial apoptotic proteins causing an increase in the ratio of Bax/Bcl2 (Bax: Bcl-2-associated X protein; Bcl-2: B-cell lymphoma 2). An insult of light to RGC-5 cells, unlike that induced by sodium azide, is inhibited by necrostatin-I and causes an activation of AlF (apoptosis-inducing factor) with alpha-fodrin being unaffected. These studies suggest that ganglion cell death caused by insults as may occur in POAG involves various cellular signaling pathways. The selected flavonoids have diverse actions in increasing cellular defense mechanisms, and in negating the effects of ischemia and specific types of oxidative stress. The results argue for the possible use of flavonoids in the treatment of POAG to slow down ganglion cell death.

by Zhang Chun. / Thesis (Ph.D.)--Chinese University of Hong Kong, 1999. / Includes bibliographical references (p. 154-197). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Mode of access: World Wide Web. / Abstract in Chinese and English.

Liu, Ke. / Thesis (Ph.D.)--Chinese University of Hong Kong, 2013. / Includes bibliographical references (leaves 175-198). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstract also in Chinese.

Ling Eva. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2004. / Includes bibliographical references (leaves 175-202). / Abstracts in English and Chinese. / Abstract --- p.i / 論文摘要 --- p.iv / Acknowledgement --- p.vi / Abbreviations Frequently Used --- p.vii / Table of Contents --- p.viii / Chapter Chapter One --- Introduction --- p.1 / Chapter Chapter Two --- Materials and Methods --- p.35 / Chapter Chapter Three --- Results --- p.64 / Chapter Chapter Four --- Discussion --- p.148 / References --- p.175 / Tables --- p.203

Ng, Tsz Kin. / Thesis (Ph.D.)--Chinese University of Hong Kong, 2010. / Includes bibliographical references (leaves 151-185). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstract also in Chinese.

Tam, Oi Sin Pancy. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2007. / Includes bibliographical references (leaves 101-128). / Abstracts in English and Chinese. / Table of Contents / Title page --- p.i / Abstract --- p.iii / 摘要 --- p.vi / Acknowledgements --- p.viii / Table of Contents --- p.ix / List of Figures --- p.xiii / List of Tables --- p.xv / Abbreviations --- p.xvii / Publications related to the work of this thesis --- p.xx / Conference Presentations related to this thesis --- p.xxi / Chapter Chapter 1 --- Introduction / Chapter 1.1 --- AMD --- p.1 / Chapter 1.2 --- Epidemiology --- p.4 / Chapter 1.3 --- Classification --- p.5 / Chapter 1.3.1 --- Dry AMD --- p.6 / Chapter 1.3.2 --- Wet/Exudative AMD --- p.9 / Chapter 1.4 --- Etiology and risk factors of AMD --- p.10 / Chapter 1.4.1 --- Gender and Ethnicity --- p.10 / Chapter 1.4.2 --- Smoking and vascular factors --- p.11 / Chapter 1.4.3 --- Genetic Factor --- p.11 / Chapter 1.5 --- Molecular Genetics of AMD --- p.12 / Chapter 1.5.1 --- Linkage studies --- p.12 / Chapter 1.5.2 --- Candidate genes search --- p.15 / Chapter 1.5.3 --- Genome-Wide Association --- p.18 / Chapter 1.5.3.1 --- Complement Factor H --- p.20 / Chapter 1.5.3.2 --- LOC387715 --- p.22 / Chapter 1.6 --- Statistical Analysis --- p.23 / Chapter 1.6.1 --- Genotyping --- p.23 / Chapter 1.6.2 --- Quality Assessment of Genetic Data --- p.24 / Chapter 1.6.3 --- Association Analysis --- p.26 / Chapter 1.6.4 --- Population Stratification --- p.26 / Chapter 1.6.5 --- Haplotype Analysis of Multiple SNPs --- p.26 / Chapter 1.6.6 --- Population Attributable Risk --- p.27 / Chapter 1.6.7 --- Interaction analysis --- p.28 / Chapter 1.7 --- Objectives --- p.28 / Chapter Chapter 2 --- Materials and Method --- p.30 / Chapter 2.1. --- Materials --- p.30 / Chapter 2.1.1. --- Proteins --- p.30 / Chapter 2.1.2. --- Chemicals --- p.30 / Chapter 2.1.3. --- Solutions and Buffers --- p.31 / Chapter 2.1.4. --- Reagents and Kits --- p.31 / Chapter 2.1.5. --- Apparatus --- p.32 / Chapter 2.1.6. --- Softwares --- p.32 / Chapter 2.2. --- Methods --- p.32 / Chapter 2.2.1. --- Study Subjects --- p.33 / Chapter 2.2.2. --- AMD atients --- p.33 / Chapter 2.2.3. --- Control Subjects --- p.34 / Chapter 2.2.4. --- DNA Extraction and Quantification --- p.34 / Chapter 2.2.5. --- Whole genome wide SNP scanning --- p.34 / Chapter 2.2.6. --- HTRA1 Genotyping --- p.38 / Chapter 2.2.6.1. --- Serial Polymerase Chain Reactions --- p.38 / Chapter 2.2.6.2. --- Cycle sequencing --- p.40 / Chapter 2.3. --- Statistical analysis --- p.40 / Chapter 2.3.1. --- Hardy-Weinberg Equilibrium Test --- p.40 / Chapter 2.3.2. --- Association Analysis: Linkage disequilibrium --- p.42 / Chapter 2.3.3. --- Haplotype Analysis --- p.43 / Chapter 2.3.4. --- Interaction Analysis --- p.43 / Chapter Chapter 3 --- Results --- p.46 / Chapter 3.1. --- Genome-wide Association Study of Exudative AMD --- p.46 / Chapter 3.1.1. --- Genotyping and Association Analysis --- p.46 / Chapter 3.1.2. --- Haplotype Analysis --- p.50 / Chapter 3.2. --- HTRA1 Genotyping --- p.57 / Chapter 3.2.1. --- Association Analysis --- p.57 / Chapter 3.2.2. --- Haplotype Analysis --- p.68 / Chapter 3.2.3. --- rsl 1200638 - Smoking Interaction --- p.68 / Chapter 3.2.4. --- rsl 1200638 - rs800292 Interaction --- p.74 / Chapter Chapter 4 --- Discussion --- p.79 / Chapter 4.1. --- Genome-wide Association Study of Exudative AMD --- p.79 / Chapter 4.1.1. --- Limitations and Concerns of Genome-Wide Association Study --- p.84 / Chapter 4.2. --- HTRA1 Genotyping --- p.85 / Chapter 4.2.1 --- Association and Haplotype Analysis --- p.85 / Chapter 4.2.2. --- HTRA1 --- p.87 / Chapter 4.2.3. --- Gene-Environment Interaction --- p.93 / Chapter 4.2.4. --- Gene-Gene Ineraction --- p.94 / Conclusions and Future Aspects --- p.97 / Electronic-Database Information --- p.100 / References --- p.101

Thesis (Ph. D.)--University of Florida, 2003. / Title from title page of source document. Includes vita. Includes bibliographical references.

Introduction: Oxygen delivery, or utilization, is a function of retinal blood flow and blood oxygen saturation. The retinal pigment epithelium (RPE), in particular, has been shown to have the highest levels of metabolic activity within the human body. Oxygen delivery is therefore of extreme importance to the maintenance of the health and integrity of the retina. Animal models presuppose that the oxygen tension in the retina is highest in the innermost layers at the level of the choriocapillaris, less in the photoreceptors and further decreases throughout the outer retinal structures. The choroid provides by far the largest component of the oxygen for consumption by the photoreceptors. A lack of oxygen stores in the inner retina therefore makes a constant supply crucial for its normal functioning. Blood flow dysfunction and subsequent hypoxia are both a feature in the pathogenesis of several major ocular diseases such as retinopathy of prematurity (ROP), age-related macular degeneration (ARMD), diabetic retinopathy (DR) and glaucoma. The development of methods to measure retinal blood flow and blood oxygen saturation is crucial to improve understanding of the patho-physiology of major ocular diseases. Purpose: The aims of this work were, firstly, to determine the least variable (range ± standard deviation) wavelength combination (610/548, 600/569 and 605/586) and subsequent ODR with the prototype HRC device. Secondly, using the ODR with the lowest measurement variability, we sought to quantify retinal blood SO2 in arterioles and venules and investigate the relationship between retinal blood SO2 and total retinal blood flow (TRBF) in response to stepwise changes in PETO2 in healthy participants. Retinal blood SO2 and TRBF were assessed using the IRIS HRC (Photon etc. Inc. Montreal, Canada) and the RTvue Doppler Fourier Domain OCT (Optovue Inc, Freemont, CA) instruments, respectively. Methods: Ten healthy participants between the ages of 23 and 37, with an average age of 28.3 years were evaluated in two descriptive cross-sectional studies. Two gas provocation protocols; hyperoxia (end-tidal oxygen; PETO2 of 100, 200, 300, 400mmHg) and hypoxia (PETO2 of 100, 80, 60, 50mmHg) were administered in a fixed sequential order. In each phase of gas provocation (via modulation of PETO2), retinal blood SO2 and TRBF measurements were acquired with the HRC and Doppler FD-OCT. The precise and repeated control of the partial end tidal pressures of oxygen (PETO2) and carbon dioxide (PETCO2) over the pre-determined phase duration, irrespective of the individuals’ respiratory rate, was made possible with the RespirAct (Thornhill Research Inc., Toronto, Canada); a sequential re-breathing gas delivery Results: In arterioles, the group range (±SD) of ODR values for baseline measurements (PETO2 of 100mmHg) was 0.169±0.061 for the 605/586 wavelength combination, 0.371±0.099 for the 600/569 wavelength combination and 0.340±0.104 for the 610/548 wavelength combination. In venules, the group range (±SD) of ODR values was 0.600±0.198 for the 605/586 wavelength combination, 0.569±0.169 for the 600/569 wavelength combination and 0.819±0.274 for the 610/548 wavelength combination. With the 605/586 combination at baseline 1 and 2 in arterioles, the group range (±SD) of ODR values was 0.607 ± 0.224 and 0.619 ± 0.158, respectively (p = 0.370), while in venules the group range (±SD) of ODR at baseline 1 and 2 was 0.289±0.750 and 0.284 ± 0.729, respectively (p = 0.714). For the 600/569 combination at baseline 1 and 2 in arterioles, the group range (±SD) of ODR values was 0.747±0.350 and 0.761±0.391, respectively (p = 0.424) while in venules the group range (±SD) of ODR at baseline 1 and 2 was 0.329±0.675 and 0.366±0.659, respectively (p = 0.372). For the 610/548 combination at baseline 1 and 2 in arterioles, the group range (±SD) of ODR values was 0.604±0.263 and 0.685±0.450, respectively (p = 0.056) while in venules, the group range (±SD) of ODR at baseline 1 and 2 was 0.292±0.746 and 0.285±1.009, respectively (p = 0.131). There was no statistical difference found between baseline ODR values (baseline 1 and 2) across all three wavelength combinations in both arterioles and venules. The mean retinal blood SO2 value at baseline in arterioles for 4 participants was 95.19% ± 31.04% and venules was 53.89% ± 17.24% (p = 0.115). There was a negative linear relationship between group retinal blood SO2 and TRBF values in the 10 participants studied, although the results of any of the 10 individuals did not show evidence of such a relationship using the described methodology. The Pearson’s correlation coefficient (r) between TRBF and SaO2 was r = -0.354 and p = 0.001 and between TRBF and SvO2 was r = - 0.295, p = 0.008 Conclusion: Of the three wavelength combinations investigated (605/586, 600/569 and 610/548), the 605/586 combination was shown to have the overall least variability. It would be unwise at this stage to adopt this wavelength combination for clinical usage, however, since it is presupposed that the 605/586 combination is also the most reliable combination to detect change in retinal blood SO2 i.e. lower variability of the 605/586 combination may be irrelevant if this combination proves to be insensitive to change in retinal blood SO2. The absolute mean ± SD retinal blood SO2 in the arterioles (SaO2) was 95.19% ± 31.04% and in the venules (SvO2) was 53.89% ± 17.24%. These values fell within the range expected and described in the literature. The magnitude of the difference between the SaO2 and SvO2 was also consistent with the literature. These findings were all appropriate for a low flow, high oxygen exchange vascular network typical of the inner retinal vascular system. Using group rather than individual data, TRBF was found in this study to relate inversely with SaO2 (r = -0.354 and p = 0.001) and SvO2 (r = – 0.295 and p=0.008), respectively. This relationship between TRBF and SaO2 and SvO2, was as expected based upon data derived primarily from animal models. This study is ground-breaking and unique, in that, it is the first study to concomitantly measure both retinal blood SO2 and TRBF in human participants. Individual data showed extensive variability and noise, thus limiting the strength of the association between TRBF and SaO2 and SvO2..

Thesis (Ph. D.)--University of Washington, 2003. / Vita. Includes bibliographical references (leaves 80-91).