ANALYSIS OF OCT IMAGES FOR LAYER SEGMENTATION AND DISEASE

Abstract

Digital fundus images are normally used for screening of retina to detect different diseases but fundus images do not provide detailed imagery of retinal layers. Optical Coherence Tomography (OCT) Imaging is a modern technique for taking cross sectional images with a very high resolution. OCT is just like ultrasound. The difference lies in method of imaging. Sound is used in ultra sound imaging whereas light is used when taking OCT images. This project deals with the analysis of OCT images and provides a prototype for detection of different retinal diseases using OCT images. The implemented system removes noise from input OCT image and extracts out all layers using different gradient and distance based operators. The project consists of a user interface to help the ophthalmologists in analysis of OCT and generates a report showing widths of all layers. The implemented system is tested on different OCT images gathered from armed forces institute of ophthalmology (AFIO).