Abstract:
Age-related Macular Degeneration (AMD) is an eye disease which affects elderly people. Cholesterol deposits in central part of retina, known as macula, damages the photoreceptors present in a particular area of eye. AMD usually effects only central vision of patient. In medical field various imaging techniques are used for diagnosis of eye diseases. Optical Coherence Tomography (OCT) is a relatively newer technique that is found to be very useful in analyzing eyes. In this research, we used OCT images to automatically detect and classify AMD. First we extract the retinal layer known as Retinal Pigment Epithelium by utilizing Graph Theory Dynamic Programing technique, after successfully enhancing the quality of OCT image by using Wiener filter. We used a unique feature set consisting of features extracted from difference signal of RPE and Inner Segment Outer Segment layer of RPE. Feature set includes approximation coefficient, Shannon’s energy, entropy and spectrum energy of the resulting difference signal. Support Vector Machine classifier was used to classify AMD affected and normal image. The developed system gives an accuracy of 95% for AMD detection.
