Abstract
Accurate in-vivo evaluation of the structural and morphological changes of the ocular structures of small animal models is essential in understanding the disease mechanisms, monitoring the disease progression, and response to therapies. Currently, ocular morphology and structure of animal models can only be examined quantitatively with histology. We present the advances for in vivo non contact three-dimensional (3D) ocular imaging of small animals with ultra high-resolution Spectral Domain Optical Coherence Tomography (SD-OCT) and quantitative information extraction using 3D segmentation of the OCT images. A SD-OCT system with two dedicated optical delivery systems for imaging the retina and anterior segment of small animals in vivo was built. An advanced 5-axis animal positioning and alignment system was developed for high throughput applications. 3D segmentation algorithm was developed for calculation of the retinal thickness map. An algorithm for automatic segmentation of the tumor boundaries and tumor volume calculation for LHβTag mouse model of retinoblastoma was also developed. The OCT system was applied in imaging mice, rats, rabbits, and raptors in the study of various ocular diseases and treatment procedures. One of the exciting applications is that the retina of Broad-winged hawk, Barred owl and Great-horned owl were imaged in vivo for the first time with high quality. The SD-OCT system accomplished the goal of non-invasive, non-contact, in vivo imaging of small animal retinal structures with high imaging quality and short imaging time (~2 minutes, acquisition time 2.7 seconds). These results make the system suitable for routine high throughput applications. Together with the segmentation algorithms, the acquired 3D data allows quantitative information extraction and provides means for precise comparison of the images acquired at different time, which make possible longitudinal studies of retinal diseases and treatment effects.