Glaucoma is an umbrella term for a number of related optic neuropathies which have the common pathology of a progressive, irreversible vision loss associated with atrophy of retinal ganglion cells. Together, the various forms of glaucoma constitute the second leading cause of vision loss in the developed world. Current therapies for the treatment of glaucoma focus on alleviating the primary risk factor, an elevation in intraocular pressure. These treatments are effective at mitigating the progression of vision loss however they cannot recover vision and do not completely halt vision loss, limiting their use as treatments. To better understand the biology underlying the loss of retinal ganglion cells in glaucoma, I have examined the role of complement in retinal ganglion cell loss. Complement is a network of cross-reacting serine proteases which form part of the humoral immune system and are primarily responsible for clearance of apoptotic cells and defence against pathogens. To understand the role played by complement in glaucoma I used an inducible model of glaucoma to establish the complement activation occurs in the glaucomatous retina. I then used the inhibitor of the classical complement cascade, C1 inhibitor to protect the dendrites and cell bodies of retinal ganglion cells and found that this protection did not prevent axonal degeneration. Using in vitro and in vivo animal models of complement deficiency I established that deficiency in complement components C3 and C6 which are further down the cascade, exacerbates damage suffered in hypertensive glaucoma. This study adds to the existing evidence that the role of complement in central nervous system degeneration is a complex, multifactorial process, with elements of the complement system being variously protective and damaging. It does, however, add hope to the prospect of developing a treatment for glaucomatous optic neuropathy based on manipulation of the complement system.

Cystic fibrosis (CF) is caused by defective function of CF Transmembrane Conductance Regulator (CFTR), an epithelial ion channel that facilitates chloride secretion. Previous research has identified impaired dark adaptation (DA) in CF, which has been attributed to concomitant vitamin A deficiency or CF-related diabetes (CFRD). However, CFTR has been localised to the retinal pigment epithelium (RPE) and it is proposed that abnormal DA could be a primary manifestation of CF. DA is similarly impaired in individuals with type 1 and 2 diabetes and is thought to be caused by retinal hypoxia as oxygen inhalation ameliorates abnormal thresholds. The aim of this thesis was to investigate DA during oxygen inhalation in CF subjects with and without CFRD to gain further insight about the aetiology of this abnormal DA. The work also aimed to examine retinal structure using optical coherence tomography (OCT) to determine the consequences of CFTR dysfunction at the RPE. Final DA thresholds were not impaired in CF subjects as a whole during the inhalation of air. However, when grouped according to diabetic status, CFRD subjects showed a significantly elevated final rod threshold which was ameliorated following oxygen inhalation. This suggests that the retina is hypoxic in CFRD subjects and that impaired DA in CF is secondary to CFRD rather than a primary manifestation of CFTR malfunction at the RPE. Contrary to the proposed hypothesis, retinal and RPE/photoreceptor layer thickness was significantly thinner in CF subjects. These results suggest that impaired CFTR function at the RPE does not directly affect retinal structure. · In conclusion, this is the first study to determine that retinal structural and functional abnormalities are not caused directly by CFTR dysfunction but are a secondary manifestation of the disease. Further research is necessary to understand the impact of these findings.

Glaucoma is a leading cause of visual disability across the world and when diagnosed the glaucoma patient will spend the rest of their life receiving treatment in managed clinical care. In the glaucoma clinic, retinal and optic nerve head (ONH) imaging can be used to help the clinician to manage patient treatment appropriately. By providing high resolution images of the optic nerve head structures and identifying changes therein related to disease onset and progression, an objective measure can be obtained as to how well or badly treatment is preventing further disease damage. This thesis contributes to the field of glaucoma progression detection by the analysis of clinical imaging data using confocal scanning laser tomography (CSLT). Primarily it is an investigation of how best to appraise and optimise current algorithms which aim to detect these glaucomatous structural changes in the optic nerve head. This is done by addressing how the performance of these methods can be best assessed in the absence of a gold standard for glaucomatous structural progression. Glaucoma expert assessment of photographs of the optic disc is the current clinical standard of assessing glaucomatous damage evident in the ONH. This is used in this thesis to act as a reference standard by which these algorithms can be compared. In addition, the statistical principles underpinning trend detection techniques are also investigated along with the performance of these techniques to detect trends in CSLT data in the presence of different types of measurement noise and image quality. A new computer model is developed and validated to simulate stable series of CSLT images, with realistic variability, which can be used to benchmark the false-positive rates of current and future progression algorithms. In conclusion, the main results reported in this thesis show that uncertainties involved in expert assessment of change in ONH photographs limits this as a reference standard for structural change in glaucoma. In addition, since stability in clinical datasets is uncertain, simulation using modelled series is shown to provide a new benchmark for comparing methods of progression detection.

The application of multifocal techniques to the visual evoked cortical potential permits objective electrophysiological mapping of the visual field. The multifocal visual evoked cortical potential (mfVECP) presents several technical challenges. Signals are small, are influenced by a number of sources of noise and waveforms vary both across the visual field and between subjects due to the complex geometry of the visual cortex. Together these factors hamper the ability to distinguish between a mfVECP response from the healthy visual pathway, and a response that is reduced or absent and is therefore representative of pathology. This thesis presents a series of methodological investigations with the aim of maximising the information available in the recorded electrophysiological response, thereby improving the performance of the mfVECP. A novel method of calculating the signal to noise ratio (SNR) of mfVECP waveform responses is introduced. A noise estimate unrelated to the response of the visual cortex to the visual stimulus is created. This is achieved by cross-correlating m-sequences which are created when the orthogonal set of m-sequences are created but are not used to control a stimulus region, with the physiological record. This metric is compared to the approach of defining noise within a delayed time window and shows good correlation. ROC analysis indicates a small improvement in the ability to distinguish between physiological waveform responses and noise. Defining the signal window as 45-250ms is recommended. Signal quality is improved by post-acquisition bandwidth filtering. A wide range of bandwidths are compared and the greatest gains are seen with a bandpass of 3 to 20Hz applied after cross-correlation. Responses evoked when stimulation is delivered using a cathode ray tube (CRT) and a liquid crystal display (LCD) projector system are compared. The mode of stimulus delivery affects the waveshape of responses. A significantly higher SNR is seen in waveforms is shown in waveforms evoked by an m=16 bit m-sequence delivered by a CRT monitor. Differences for shorter m-sequences were not statistically significant. The area of the visual field which can usefully be tested is investigated by increasing the field of view of stimulation from 20° to 40° of radius in 10° increments. A field of view of 30° of radius is shown to provide stimulation of as much of the visual field as possible without losing signal quality. Stimulation rates of 12.5 to 75Hz are compared. Slowing the stimulation rate produced increases waveform amplitudes, latencies and SNR values. The best performance was achieved with 25Hz stimulation. It is shown that a six-minute recording stimulated at 25Hz is superior to an eight-minute, 75Hz acquisition. An electrophysiology system capable of providing multifocal stimulation, synchronising with the acquisition of data from a large number of electrodes and performing cross-correlation has been created. This is a powerful system which permits the interrogation of the dipoles evoked within the complex geometry of the visual cortex from a very large number of orientations, which will improve detection ability. The system has been used to compare the performance of 16 monopolar recording channels in detecting responses to stimulation throughout the visual field. A selection of four electrodes which maximise the available information throughout the visual field has been made. It is shown that a several combinations of four electrodes provide good responses throughout the visual field, but that it is important to have them distributed on either hemisphere and above and below Oz. A series of investigations have indicated methods of maximising the available information in mfVECP recordings and progress the technique towards becoming a robust clinical tool. A powerful multichannel multifocal electrophysiology system has been created, with the ability to simultaneously acquire data from a very large number of bipolar recording channels and thereby detect many small dipole responses to stimulation of many small areas of the visual field. This will be an invaluable tool in future investigations. Performance has been shown to improve when the presence or absence of a waveform is determined by a novel SNR metric, when data is filtered post-acquisition through a 3-20Hz bandpass after cross-correlation and when a CRT is used to deliver the stimulus. The field of view of stimulation can usefully be extended to a radius of 30° when a 60-region dartboard pattern is employed. Performance can be enhanced at the same time as acquisition time is reduced by 25%, by the use of a 25Hz rate of stimulation instead of the frequently employed rate of 75Hz.

Bacterial keratitis and bacterial endophthalmitis are two of the most devastating sight threatening eye infections. It is currently unclear whether the organisms isolated from these infections represent specialised members of these species or whether all strains are equally likely to cause infection. One method of differentiating strains genotypically is by a typing technique known as multilocus sequence typing (MLST). Using this technique, a better understanding of the molecular epidemiology of eye infections can be achieved. Propionibacterium acnes (P. acnes) is an important anaerobic organism causing several types of eye infections. Although it is now beginning to be recognised as a serious opportunistic pathogen, few studies have been done to investigate the population biology of P. acnes at the molecular level. Our continuing inability to distinguish between strains of P. acnes means that we still do not fully understand how antibiotic-resistant strains spread, nor whether certain strains, or clonal complexes, of P. acnes are associated with certain infections. These are key issues that can now be understood with our development of an MLST system for P. acnes. A diverse culture collection of 125 P. acnes isolates have been analysed using the MLST scheme developed. Sequence analysis shows that there are phylogenetically distinct groups within P. acnes and identified a novel cluster not previously described. Analysis of recombination using several methods suggests that frequent recombination occurs within these subgroups. There appears to be no association between these subgroups and clinical manifestation of P. acnes infection or geographical location. The P. acnes MLST scheme was validated against 16S rRNA gene and complete recA gene typing as well as immunofluorescence microscopy (IFM) and random amplification of polymorphic DNA (RAPD) analysis. P. acnes is a slow growing organism and is difficult to culture from ocular samples. A real-time PCR assay was developed in order to overcome the low culture positive rate and delay associated with conventional culture methods. Primers targeting one of the seven housekeeping genes used in the MLST scheme (gmk), specific for P. acnes were selected. The real-time PCR assay was both specific to P. acnes and highly sensitive. Pseudomonas aeruginosa (P. aeruginosa) is another important organism in the development of serious eye infections, and is the commonest cause of contact lens related microbial keratitis. Infections with this organism can lead to rapid deterioration of vision and possible blindness. A previously developed MLST scheme has been applied to 117 eye isolates from around the UK and China. This typing data was compared to 166 P. aeruginosa isolates from other clinical and environmental sources. Overall, MLST data supports previous findings that P. aeruginosa has a nonclonal population with epidemic clones. Sequence analysis showed that eye isolates do not cluster away from isolates from other clinical infection sites.

Infantile Nystagmus Syndrome (INS) is defined as a constant involuntary movement of the eyes, affecting 0.12% of the population. Previous research shows that eye movements in patients with nystagmus increase in stressful conditions, and they report that their vision gets worse. However, there have been no definitive conclusions as to the effect of stress on visual acuity (VA). The aim of the studies described here was to assess visual function, including VA, during periods of stress. The results showed that there was no difference in VA measured with horizontally and vertically oriented Landolt C’s, but, in agreement with published research, VA was found to be poorer when using vertically oriented gratings as compared to horizontal gratings. Using a Trans-Cutaneous Electrical Nerve Stimulation (TENS) machine, an effective clinical stressor, we confirmed that the intensity of nystagmus increased when under stress; however VA, as measured using Landolt Cs, was not affected. Patient response time also increased during stressful periods and was significantly longer in INS than in control subjects. Using a questionnaire, we identified the most stressful situations for patients with nystagmus as being: “finding a person in a crowd” and “crossing a road in heavy traffic”. Under stress, rather than vision becoming blurred, patients with nystagmus reported that they “take longer to see things” and “have difficulty with seeing facial features and small detail”. The results reported here have important implications for patients with nystagmus in the real world, strongly suggesting that, although maximum acuity is unaffected by stress, more time should be allowed for tasks at both work and school. Further research is required to fully understand the changes in other aspects of visual function with stress.

Recent studies have indicated that avian magnetoreception is based on vision, being both wavelength and luminance sensitive. Using this novel behavioural assay, the ability to respond to changes in magnetic field conditions was tested in the dark and the light. Darkness was found to have an adverse impact on birds' abilities to perceive magnetic field changes. Magnetoreception as measured by reactivity was virtually eliminated in the dark as compared with the same birds' behaviour in normal light conditions.

The current understanding of anterior eye shape in humans is limited due to available technology and its accessibility. Accurate curvature metrics of specific areas of the peripheral cornea, corneo-limbal junction and anterior sclera have remained obscured by the limits of the palpebral aperture, since the upper and lower eyelids cover most of the vertical aspect. This thesis starts by comparing the ‘gold standard’ keratometry measurements to commonly used topographic systems. Keratometric analogues were found to be significantly different and in addition provided spurious vertical anterior ocular surface (AOS) profiles. These findings revealed a need to establish an accurate model. Magnetic resonance imaging (MRI) potentially offers the best opportunity to image the entire AOS structure. However, preliminary studies in this thesis demonstrated that the use of a 3-Tesla MRI scanner was unable to obtain sufficiently resolute data to meet requirements. As an alternative, ocular impression taking techniques were adopted during the remainder of this work to acquire the AOS data. Eye casts from impression moulds were scanned using active laser triangulation and virtual 3-dimensional surfaces rendered. Further investigations defined the most suitable material for impression taking and the amount of deformation of the AOS caused by the procedure. The ocular impression casting and scanning process was examined for accuracy and reliability. This protocol was used to sample a population of normal white European eyes in order to establish a database and define wide-field AOS variability. Volumetric and 2-dimensional topographic profiles were extracted from the digital 3-dimensional representation obtained, allowing for the analysis of point-to-point curvature differences. For the first time, the entire AOS shape has been defined with known accuracy. In addition, effects of myopic refractive error and gender are presented. This data is of potential importance to ophthalmic surgeons, ocularists, contact lens practitioners, vision scientists and researchers, in the form of a digital archive of normal white European wide-field AOS topography as a reference source.

Age-related macular degeneration (AMD) is the principle cause of visual loss and blindness in the developed world. As new treatments and therapies are developed the need to better diagnose and then monitor outcomes of treatment has become more important. This thesis evaluates both structural and functional changes that occur in the early stage of AMD, known as age-related maculopathy (ARM), with the aim of determining their diagnostic potential. This thesis also explores the relationship between structural and functional parameters. Twenty four participants with ARM and 26 control participants were recruited. Retinal function was probed using four focal electroretinography (ERG) techniques: the focal cone ERG, focal flicker ERG, ERG photostress test and focal rod ERG. Long wavelength optical coherence tomography (OCT) was used to assess retinal structure, specifically retinal, choroidal and four intra-retinal layer thicknesses at 21 macular locations. These techniques were initially developed and optimised for the detection of AMD related changes. The ability of each parameter to diagnose ARM was assessed. Correlation and linear regression analyses were carried out to identify any relationships between retinal structure and function in healthy controls. Retinal thickness was reduced in participants with ARM at parafoveal locations (~2° eccentricity), but choroid thickness was unaffected. Diagnostically, focal ERG parameters provided better sensitivity and specificity to ARM than OCT measures, with the ERG photostress test providing the best diagnostic potential. No strong relationships were shown between any ERG parameter and any retinal or choroidal layer volume in control participants. Three ERG parameters were shown to be related to specific retinal features of ARM, but the strongest associations were between ERG photostress test recovery and focal cone ERG b-wave implicit time and a diagnosis of wet AMD in the contralateral eye. In conclusion the structural and functional parameters assessed appeared to provide independent information regarding disease status and severity. ERG parameters showed better diagnostic potential than OCT measures. The single most diagnostic parameter was the recovery time constant of the ERG photostress test.

The aim of this PhD thesis was to develop segmentation algorithm adapted and optimized to retinal OCT data that will provide objective 3D layer thickness which might be used to improve diagnosis and monitoring of retinal pathologies. Additionally, a 3D stack registration method was produced by modifying an existing algorithm. A related project was to develop a pre-apoptotic retinal monitoring based on the changes in texture parameters of the OCT scans in order to enable treatment before the changes become irreversible; apoptosis refers to the programmed cell death that can occur in retinal tissue and lead to blindness. These issues can be critical for the examination of tissues within the central nervous system. A novel statistical model for segmentation has been created and successfully applied to a large data set. A broad range of future research possibilities into advanced pathologies has been created by the results obtained. A separate model has been created for choroid segmentation located deep in retina, as the appearance of choroid is very different from the top retinal layers. Choroid thickness and structure is an important index of various pathologies (diabetes etc.). As part of the pre-apoptotic monitoring project it was shown that an increase in proportion of apoptotic cells in vitro can be accurately quantified. Moreover, the data obtained indicates a similar increase in neuronal scatter in retinal explants following axotomy (removal of retinas from the eye), suggesting that UHR-OCT can be a novel non-invasive technique for the in vivo assessment of neuronal health. Additionally, an independent project within the computer science department in collaboration with the school of psychology has been successfully carried out, improving analysis of facial dynamics and behaviour transfer between individuals. Also, important improvements to a general signal processing algorithm, dynamic time warping (DTW), have been made, allowing potential application in a broad signal processing field.