Abstract

We present our progress on image-guided (sensor-less) adaptive optics for multi-modal imaging of retinal tissue using high numerical aperture OCT and confocal microscopy using a custom developed instrument. Images of ex vivo tissues are compared to data acquired in vivo.

Introduction Imaging the retina non-invasively is a valuable tool for vision research. Optical Coherence Tomography (OCT) is highly valued for the ability to provide structural images of the retinal layers, but lacks molecular contrast. Complementary imaging modalities are required to identify specific cells and tissues of interest. Fluorescence is a convenient source of contrast in the retina due to the presence of endogenous fluorophores and the ease of introducing extrinsic fluorophores. As an example, fluorescein is commonly used clinically for labeling the retinal vasculature. In preclinical imaging, the relevant cells can be labelled with a fluorophore such as Green Fluorescent Protein (GFP), complementing autofluorescence from layers such as the Retinal Pigment Epithelium (RPE). The RPE is of particular interest due to its connection to diseases such as Age Related Macular Degeneration (AMD). AMD is the leading cause of vision loss for adults in the UK, affecting more than 500,000 people with a total annual economic cost of £2.6 billion [1]. Progress in the imaging of this layer would make great strides in the development of novel therapies for such diseases. We are performing a comparative study of imaging with OCT and fluorescently labeled targets with high resolution. When imaging the retina through an ophthalmoscope, and given that the focal length of the eye is fixed, the numerical aperture (NA) is maximized by filling the pupil, which introduces monochromatic aberrations. This severely distorts the focal spot, and therefore dims the image significantly. To improve the images in the presence of aberrations, we use adaptive optics (AO) to obtain a near-diffraction limited focused spot at the retina. We employ a sensor-less adaptive optics (SAO) approach via an image-based hill-climbing algorithm to determine which aberration corrections result in the sharpest image. We have previously reported on small animal retinal imaging systems that can perform image-guided aberration corrections using SAO (see for example [2] and [3]. In this report, we present on our progress on evaluating the performance of our multimodal imaging system combining AO, OCT and scanning laser confocal reflectance and fluorescence on ex vivo retina samples.

© 2023 SPIE