Visible Optical Coherence Tomography based Multimodal Imaging for Quantification of Retinal Lipofuscin

Abstract

Listen

Translate article

Retinal degeneration is the leading cause of irreversible low vision and blindness in the world, that describes conditions characterized by progressive loss of photoreceptors. Retinal Pigment Epithelium (RPE) is located under photoreceptors’ outer segments and plays an important role in the maintenance of photoreceptors by completing the visual cycle and phagocytosis of shed photoreceptor outer segments. Lipofuscin, a byproduct of the visual cycle, is a nondegradable compound that accumulates in the RPE cells and eventually damages the RPE cells and inevitably causes photoreceptor degeneration. Lipofuscin is the major cause of fundus fluorescence that can be detected by Fundus Autofluorescent (FAF) imaging systems. Reliable and quantified FAF values are necessary for lipofuscin quantification which can be a significant tool in the diagnosis of retinal degenerative disease in early stages and provide a better opportunity for treatment before the loss of vision stage. However, FAF is attenuated by the ocular media prior to the RPE, including cornea, lens, vitreous body, retinal layers in front of the RPE, and the melanin granules within the RPE cells that migrate to the apical region upon light exposure. This attenuation varies among people and for an individual over time and cannot be measured directly, thus hurdles measurement of the true FAF values. Further, differences in acquisition systems such as illumination power and detector sensitivity, directly affect the measured FAF. This issue has been addressed by implementing a reference target in the FAF imaging system. Normalizing the FAF signal to that of the target eliminates the dependency on the acquisition parameters. However, the issue of pre-RPE and RPE melanin attenuation remains unresolved. Further, the fluorescence characteristics of the commercially available fluorescent reference are quite different than retinal lipofuscin that challenges the quantification of the absolute amount of lipofuscin in the RPE. In this dissertation, we propose a new multimodal imaging system based on visible-light optical coherence tomography (VIS-OCT) that provides a three-dimensional image. The technology simultaneously acquires VIS-OCT and FAF with a single broadband visible light source. Since both images are originated from the same group of photons and travel through the same ocular media at the same time, the attenuation factor is similar in both modalities. Therefore, by normalizing FAF by VIS-OCT of the RPE layer, the attenuation of the pre_RPE media can be eliminated. Further, we implemented two reference targets to quantify VIS-OCT and FAF and eliminate the dependency on acquisition parameters. These references were