الفهرس 
Chapter 1 IntroductionOnly 14 pages are availabe for public view
 |  | from | 119 |  |  |
| | | from | 119 | | |
Abstract
According to the statistics about glaucoma, Glaucoma is the most frequent disease in worldwide; with nearly 1.67 million new cases diagnosed .Early detection of glaucoma increases the prognosis and hence increases the rate of healthy people. Optical Coherence Tomography is still the widely used eye-screening tool for detectingdiseases. Many computer vision techniques have been proposed to increase the accuracy of detecting eye diseases such as a Computer Aided Detection or Diagnosis (CAD), which consists of five stages. The first stage is the data acquisition, the second is the preprocessing stage, the third is the segmentation stage, and the fourth is the feature extraction and finally the classification stage. In this thesis, we dell with OCT images which obtained from Al-Hayaa advanced radiology center. The entire method presented in these chapters was implemented using MATLAB 2015, and makes extensive use of the Image Processing Toolbox. These images have a low contrast and suffer from noise so we proposed algorithm for the OCT images, whichconsist of preprocessing stage to make better visualization of images and noise removal stage. To remove the noise from image by using different filter such as median, diffusion and wiener filter depend on PSNR and MSE. We found that the best filter for all types of noise especially speckle noise is diffusion filter due to high PSNR and low MSE. In the next stage of CAD algorithm is a segmentation of retina layers which have been done with differentmethods .To select the best method of segmentation we determine the thickness of retina layer. After segmentation of the retina layer, the features of the images have been enteredto the classifier for the classification process. Image classification is a most important step for image analysis. Classification helps the doctors to treat the patient according to the diseases. Classification is a computational procedure that separates the images into groups according to their features that extracted. The output of the NN is normal or abnormal. In our thesis, we used neural network (NN) as a classifier. Classification is done on the base of parameterThesis objectives : Develop an algorithm to improve the low contrast of OCT image. Develop software programs for filtering the noises that found in the images using the selected filter (diffusion filter) using MATLAB program. Develop software programs forretina layer segmentation using different method using MATLAB. Develop software programs for image classification and training our output data to neural network to identify thickness of layer.