Over the past few decades, functional imaging has received increasing attention from both research scientists and clinical practitioners. Compared to conventional imaging technologies, functional imaging modalities can provide the variations in chemical composition, absorption, blood flow and metabolism of biological tissue in addition to visualization of its structure. Information regarding these functional variations not only sheds light upon mechanisms of fundamental physiological processes, but also facilitates early detection of diseases. This dissertation seeks to develop functional imaging modalities for eye imaging, which, by providing enriched understanding of retinal pathophysiology, can ultimately lead to improved management of ocular diseases.

A number of well-established technologies can provide functional imaging of biological tissue. For example, positron emission tomography (PET) reveals tissue metabolism by imaging regional glucose uptake; Functional magnetic resonance imaging (fMRI) and computed tomography (CT) perfusion imaging measure neural activity in brain by detecting changes in blood flow. However, with half-millimeter resolution, these imaging modalities cannot resolve the fine information in the retina. The retina is a light-sensitive layer of tissue and a part of central nervous system. It converts the light into nerve impulses and transmits the signal to visual cortex. Though containing several sub-layers with different functionalities in the visual preprocessing cascade, the total thickness of retina is only hundreds of micrometers. Evolved as the optical sensor of human body, the eye provides the light with unhindered access to retina through cornea, lens, aqueous humor and vitreous humor, which makes optical imaging suitable for retinal examination. In addition, optical imaging, in general, has micrometer-level resolution which is sufficient to resolve retinal structure. Although strong optical absorption and scattering of biological tissue diffuse the probing light beyond one-millimeter penetration depth and limit the application of optical imaging to most internal human organs, imaging the entire retina is well within its capabilities. Therefore, optical imaging is ideal for the examination of the retina.

Existing clinical retinal imaging modalities, such as fundus photography, scanning laser ophthalmoscopy (SLO) and conventional optical coherence tomography (OCT) rely on structural variation of the retina to detect eye diseases. However, there is growing evidence suggesting that functional aberration may precede structural abnormalities in the leading blinding diseases. For example, loss of melanin in the retinal pigment epithelium (RPE) is both a risk factor and early symptom of age-related macular degeneration (AMD) and measurement of optical absorption and melanin concentration can potentially lead to early detection and better monitoring of the disease. As another example, growing evidence suggests that in diabetic retinopathy (DR), the metabolic changes may reflect onset of the disease long before vascular structure aberration, and evaluation of retinal metabolism can bring medical intervention to an earlier stage.

In this dissertation, functional photoacoustic microscopy (PAM) is used to measure the melanin concentration in the RPE. PAM combines optical illumination with acoustic detection and is specific to optical absorption, or energy deposition, which is proportional to the concentration of optical absorbers. A Monte Carlo simulation is first performed to study the optical energy deposition in the retina and the generation of the photoacoustic (PA) signal. An intrinsic calibration scheme using PA signals of retinal blood vessel as reference is proposed to improve consistency across different measurements, different subjects, and different imaging systems. The simulation suggests that, though not required, a fine axial resolution that can distinguish RPE from underlying choroid, which also generates PA signal due to optical absorption, is preferred to increase the measurement sensitivity. Following the simulation, a high axial resolution PAM, enabled by a broadband ultrasonic detector based on a micro-ring resonator (MRR), is developed to measure the melanin concentration in ex vivo eye samples from pig and human. The exceptional 2.12-microm axial resolution allows separation of RPE from choroid and the melanin concentrations in both layers are quantified.

In addition to developing PAM for functional imaging of RPE melanin, this dissertation also represents a pioneering study that translates functional visible-light optical coherence tomography (vis-OCT) from preliminary laboratory investigation to clinical applications. Instead of using conventional near-infrared (NIR) light sources, vis-OCT exploits the visible band of the electromagnetic spectrum, which not only improves axial resolution but also provides additional spectroscopic contrast to evaluate oxygen saturation of blood vessel and tissue metabolism. The first vis-OCT ophthalmoscope is developed for in vivo human eye imaging. Engineering design, and user friendliness of the initial prototype system are optimized continuously, which leads to improved imaging quality and provides a platform for retinal oximetry in human subjects.

Besides developing functional imaging platforms on the system level, efforts are also made to improve the imaging performance from algorithm and light source perspectives. (Abstract shortened by ProQuest.).