Myeong Jin Ju

Preclinical studies provide valuable insights into disease mechanisms, potential treatments, and therapeutic strategies before clinical trials in the studies of retinal diseases. Mouse, one of the most widely used small animals for preclinical studies, possess great advantages as experimental animals based on their easy accessibility and availability which make them an ideal animal with a broad range of pathological models. Optical coherence tomography (OCT) allows non-invasive in vivo imaging that generates high-resolution 3D images and has been successfully applied in ophthalmology. Also, the recent progress in the development of functional extensions of OCT allowed the generation of multiple contrasts providing information-rich images. This thesis focuses on the development and verification of a polarization-diversity OCT (PD-OCT) system tailored for the mouse retina to acquire high-resolution multi-contrast images. The implementation of a sensorless adaptive optics technique and a multiple spectrometer calibration method ensures the capture of retinal images with an advanced resolution while precisely detecting melanin contrast. Preclinical application of PD-OCT demonstrated its efficacy in visualizing fine retinal features in multiple contrasts with the help of the 3D image registration technique.

View record

Optical coherence tomography (OCT) is a non-invasive optical imaging technique to capture in-vivo, three-dimensional (3D) image data with micrometer resolution. OCT has become a valuable tool for ophthalmologists in detecting and monitoring eye diseases like age-related macular degeneration (AMD), glaucoma, and diabetic retinopathy (DR) by providing valuable information for diagnosis and determining the best course of treatment. OCT angiography (OCTA) is a functional extension of OCT which allows for non-invasive visualization of blood flow in the retina and choroid. OCTA provides detailed information about microvasculature, which is not possible with traditional angiography techniques. However, OCT and OCTA images are susceptible to degradation from motion artifacts induced by the involuntary human subjects’ head and eye movement. Multi-volume acquisition and averaging is a software-based approach to correct the motion artifacts by registering several motion-corrupted images to a single reference image or multiple reference images in a post-processing step. Nonetheless, most software-based algorithms have been more specifically designed for a specific system with a limited field-of-view (FOV) range. In this study, we implemented a robust 3D algorithm for the precise registration of multiple volumes to the coordinates of one volume chosen as the reference volume. Tests were carried out on various datasets obtained from different types of custom-built and commercially available systems to determine the reliability of the proposed algorithm. A two-dimensional (2D) version of this method is then applied to investigate retinal perfusion heterogeneity, which may enhance our understanding of retinal diseases such as glaucoma.

View record