Search

Amblyopia is the leading cause of monocular visual impairment in children. Therapy for amblyopia is extremely beneficial in some children but ineffective in others. It is critical that the reasons for this discrepancy are understood. Emerging evidence indicates that current clinical protocols for the diagnosis of amblyopia may not be sufficiently sensitive in identifying individuals who, on more detailed examination, exhibit subtle structural defects of the eye. Presently, the magnitude of this problem is unknown. The aim of this study was to establish the prevalence of subtle retinal/optic nerve head defects in eyes diagnosed with amblyopia, to distinguish between possible explanations for the origin of such defects and to investigate the relationship between quantitative measures of retinal structure, retinal nerve fibre layer thickness and optic nerve head dimensions. Using the imaging technique of Optical Coherence Tomography (OCT) retinal structure has been investigated in detail, following the visual pathway across the retina from the fovea, via the paramacular bundle to the optic disc, where peripapillary retinal nerve fibre thickness has been imaged and subjected to detailed measures along with optic disc size and shape. The study formed two phases, the first imaging the eyes of visually normal adults and children, comparing them to amblyopes, both adults and children who had completed their treatment. The second phase, a longitudinal study, investigated retinal structure of amblyopic children undertaking occlusion therapy for the first time. By relating pre-therapy quantitative measures to the visual outcome the second phase of the study aimed to examine whether OCT imaging could identify children achieving a poor final outcome. The results show a clear picture of inter-ocular symmetry structure in all individuals, visually normal and amblyopic. Optic disc characteristics revealed no structural abnormalities in amblyopes, in any of the measured parameters, nor was there any association between the level of visual acuity and the measured structure. At the fovea differences were shown to occur in the presence of amblyopia, with thickening of the fovea and reduction of the foveal pit depth. The structural changes were found to be both bilateral and symmetrical with the fellow eye also affected. In the longitudinal phase of the study these changes were demonstrated to a greater extent in children who 'failed' to respond to treatment. This bilateral, symmetrical structural change found at the fovea, which has not been previously reported, cannot therefore be the primary cause of the visual loss which has been diagnosed as amblyopia.

The desire to produce artificial vision systems which behave in an intelligent, humanlike way or which can autonomously and automatically perform tasks currently only performed by humans has been a goal of Artificial Intelligence research for many decades. Until recently much of the research concentrated on extracting visual representations of objects from single, static scenes. The last decade has seen an increase in interest concerning mobile robotics for navigation, planning and autonomous control as well as for the interpretation of events in real, dynamic scenes. Presented in this thesis is research on artificial vision systems from two different, but both necessary, standpoints. One concerns low-level vision-based behaviour of object tracking based upon a naturalistic theory of perception and behaviour within living systems. The other takes a more application and engineering based approach and its goal is to address high-level scene interpretation and control of processing resources. Numerous experiments are presented to demonstrate the various issues. The two main experiments, corresponding to the two research streams, are a system which is able to fixate complex multi-coloured objects and a fully integrated vision system for predicting and following, with a mobile sensor, events in a dynamic scene.

The Henle fiber layer (HFL) is comprised of cone photoreceptor axons in the central fovea which undergo lateral displacement along with the inner retinal layers throughout early development to give rise to the foveal pit. The close relationship between the photoreceptor axons and the cone photoreceptors within the fovea allows the HFL to be studied and imaged as an indirect measure of photoreceptor viability. Due to the optical transparency of the HFL, traditional reflectance based imaging modalities, such as optical coherence tomography (OCT), have struggled to accurately delineate this retinal layer from the adjacent outer nuclear layer. By employing directional OCT imaging, which utilizes offset of the imaging beam, maximization of the primary scattering angle of the HFL can be achieved allowing the tissue to be more readily visualized and quantified in thickness. Additionally, the HFL is one of various ocular tissues which exhibit form birefringence attributed to the production and organization of microtubules within the axonal cell walls. This birefringence acts as a potential biomarker for density and integrity changes to the foveal cone photoreceptors and can be measured and quantified using polarization sensitive technology such as scanning laser polarimetry (SLP). In this study, we examine two optical properties associated with the Henle fiber layer using directional OCT and SLP to compare the HFL thickness measurements obtained using these two techniques.Directional OCT imaging was performed using the Heidelberg Spectralis spectral domain OCT in both the horizontal and vertical meridians to achieve measurements for the superior, inferior, nasal and temporal retina. Manual segmentation using Adobe Photoshop was performed and thickness values for the four meridians were averaged. Henle fiber layer phase retardation was measured using the GDx scanning laser polarimeter which generated a characteristic and expected macular cross pattern due to the radial symmetry of the tissue. Custom curve fitting analysis was used to generate intensity profiles in the phase retardation maps. Henle fiber layer thickness values and the amplitude of the phase retardation signal were plotted against each other at 0.25 degree intervals from 0 to 3 degrees eccentricity. The average thickness profiles and phase retardation amplitudes correlated closely at the eccentricities evaluated (r2 = 0.981). Overall, HFL thickness and phase retardation were minimal centrally, increased until reaching a maximum at 2 degrees eccentricity, and declined at greater eccentricities. In this study, we measured HFL thickness using directional OCT imaging and found a significant correlation with macular phase retardation distribution using SLP in normal subjects. The strong correlation between phase retardation distribution and HFL thickness in the central macula provides supporting evidence that the macular phase retardation signal results from the birefringent HFL in healthy normal subjects. Understanding this correlation becomes important, as the combined use of these unique, but complimentary techniques is potentially an optical signature for density and integrity changes to the photoreceptor axons in the central macula in normal aging and disease.

The purpose of this study is to determine whether a circular shape adequately describes the shape of the foveal pit, and if so, over what distance does this hold true. Another purpose of this study is to evaluate meridional symmetry within a normal fovea and compare radius of curvature between the horizontal and vertical meridians. The radius of curvature will also be compared to gender, axial length, and age. Previous studies have evaluated foveal pit shape and radius of curvature but using different techniques such as reflectometry, difference of Gaussians, and thickness measurements.1-4 Various studies have compared foveal characteristics to gender, axial length, and age.1-4 Many studies have also evaluated the underlying cellular organization and structure that define the fovea.5-9 The present study uses a novel curve fitting technique that is more direct and provides better visualization of the foveal curvature in more than one meridian. 22 normal subjects were imaged using the spectral domain optical coherence tomography (SDOCT, Heidelberg Engineering, Carlsbad, CA). All reported normal ocular health, and visual acuity was measured as clinically normal. Axial length measurements were also recorded. Twelve radial scans centered at the fovea were acquired from each eye. The images were exported into Photoshop (Adobe, San Jose, CA) and manually segmented. The line representing the internal limiting membrane was used for analysis in the custom curve fitting software (Matlab, Mathworks, Natick, MA). The circle was fit to the central 500 microns in a randomly selected eye of each of the 22 subjects. Linear regression was used to compare radius of curvature to axial length and age, and a paired t-test was used to compare radius of curvature between males and females. The circle was also fit to the central 500, 1000, and 1500 microns in the right eye of 10 of the 22 subjects. Paired t-tests were used to compare the horizontal and vertical meridian at each of the three locations in these 10 subjects. There was no significant association between radius of curvature in the horizontal or vertical meridian and axial length (R2=0.0378, R2=0.0264) or age (R2=0.0125, R2=0.0005) and no significant difference when compared between males and females (p=0.602, p=0.540). A circle fit to the central 500 microns demonstrated a good representation of foveal pit curvature in all subjects. The fit of the circle decreased with increased eccentricity at 1000 microns and even more so at 1500 microns in both meridians. The difference in radius of curvature between the horizontal and vertical meridians was significant at 1000 microns (p

Abstract: We designed independent perceptual video compression that can work with all the available video coding standards. We used fovea as the Human Visual System (HVS) properties in this thesis. In this thesis, first, we describe and compare two available foveation techniques used for images. After finding the pros and cons of the techniques, we improve one of the single-fovea techniques for video compression so that it preserves the quality of videos better compared to conventional video coding. Finally, we compare our technique with four foveation-based objective metrics for nine different video contents and different parameters, and in order to improve the evaluation assessment, we did two subjective tests to hear the users opinion about the video results.

The purpose of this research was to investigate detect and monitor laser-tissue interactions at threshold and potentially sub-threshold levels of injury. High resolution imaging modalities can provide a deeper understanding of candidate biomarkers disease and injury at the molecular, cellular, and tissue-levels which can be used to identify and diagnose early stages disease and damage. In addition, multi-scale and multi-modal imaging have also been used to identify inherent biomarkers of retinal disease and injury. Monitoring tissue changes can be mapped back to biological changes at the cellular and sub-cellular level. Diseases often alter tissue on the ultra-structural level yet retinal clinical diagnosis often monitor changes in tissue at the organ level. If injury and disease is detected and diagnosed during an “early” stage of development, treatments and drug interventions may prevent further spread of the pathology. Non-invasive imaging is expected to be a valuable tool for in vivo medical research as well as for the diagnosis and management of disease. In addition to developing new imaging tools and techniques to image the retina, the identification of inherent biomarkers of disease and health using diagnostic methods are almost equally as important. Using the inherent optical properties of retinal tissue, we can non- invasively quantify differences in the absorption and reflection of light to gauge the risk for visual disability or worse yet irreversible vision loss as a result of retinal disease and chronic light exposure. The research presented with in this dissertation is three separate studies aimed at identifying light injury and potential biomarkers indicating the risk of light mediated development of disease.

In order to study the limits of spatial vision in normal human subjects, it is important to look at and near the fovea. The fovea is the specialized part of the retina, the light-sensitive multi-layered neural tissue that lines the inner surface of the human eye, where the cone photoreceptors are smallest (approximately 2.5 microns or 0.5 arcmin) and cone density reaches a peak. In addition, there is a 1:1 mapping from the photoreceptors to the brain in this central region of the retina. As a result, the best spatial sampling is achieved in the fovea and it is the retinal location used for acuity and spatial vision tasks. However, vision is typically limited by the blur induced by the normal optics of the eye and clinical tests of foveal vision and foveal imaging are both limited due to the blur. As a result, it is unclear what the perceptual benefit of extremely high cone density is. Cutting-edge imaging technology, specifically Adaptive Optics Scanning Laser Ophthalmoscopy (AOSLO), can be utilized to remove this blur, zoom in, and as a result visualize individual cone photoreceptors throughout the central fovea. This imaging combined with simultaneous image stabilization and targeted stimulus delivery expands our understanding of both the anatomical structure of the fovea on a microscopic scale and the placement of stimuli within this retinal area during visual tasks. The final step is to investigate the role of temporal variables in spatial vision tasks since the eye is in constant motion even during steady fixation. In order to learn more about the fovea, it becomes important to study the effect of this motion on spatial vision tasks. This dissertation steps through many of these considerations, starting with a model of the foveal cone mosaic imaged with AOSLO. We then use this high resolution imaging to compare anatomical and functional markers of the center of the normal human fovea. Finally, we investigate the role of natural and manipulated fixational eye movements in foveal vision, specifically looking at a motion detection task, contrast sensitivity, and image fading.

Head-bobbing is the fore-aft movement of the head exhibited by some birds during terrestrial locomotion. It is primarily considered to be a response to enhance vision. This has led some researchers to hypothesize that head-bobbing should be found in birds that are visual foragers and may be correlated with the morphology of the retina. In contrast, other researchers suggest that head-bobbing is mechanically linked to the locomotor system and that its visual functions are secondarily adapted. This dissertation explored the mechanics of terrestrial locomotion and head-bobbing of birds in both the lab and field. In the lab, the kinetics and kinematics of terrestrial locomotion in the Elegant Crested Tinamous Eudromia elegans were analyzed using high-speed videography and ground reaction forces. In the field, the kinematics of the terrestrial locomotion of charadriiform birds were analyzed using high-speed videography. These biodynamic studies found that birds have two distinct locomotor transitions. The first transition occurs as they move from vaulting mechanics to bouncing mechanics, and the second transition occurs when they incorporate an aerial phase during running. Thus, many birds use grounded running during intermediate speeds. Also, this study found in general that head and neck movements are coordinated with limb movements during terrestrial locomotion in charadriiform birds, but were not coordinated in the Elegant-crested Tinamou. Thus, the coordination is neither perfect nor obligatory. Additionally, the evolutionary histories of retinal morphology, head-bobbing and foraging type in birds were investigated through a comprehensive review of avian literature. The data were compiled and mapped onto a composite avian phylogeny. Then, the correlation of characters was analyzed using pairwise comparisons. This study found that a nasal unifoveate retina with a band-shaped area centralis, non-head-bobbing and visual foraging appear to be the ancestral character states for birds. Additionally, there were no significant correlations between head movement, retinal pattern or foraging mode. Although some general trends were observed, most were within clades and, thus, a result of the independent evolutionary history of head movement, retinal pattern or foraging mode.