الفهرس | Only 14 pages are availabe for public view |
Abstract Automatic detection of maculpathy disease is a very important step to achieve high-accuracy results for the early discovery of diseases and help ophthalmologists to treat patients. Manual detection of diabetic maculpathy is time consuming and it needs much effort from ophthalmologists Retinopathy is an eye disease for diabetics. Retinopathy can occur with all types of diabetes and can lead to lack of vision if leaved untreated. Diabetic maculopathy results from the retinopathy disease. Maculopathy is damage to the macula. Macula is a sensitive part of the retina, which is used for central vision and reading. It is near the retina center. Exudates are one of the common visible signs of diabetic maculpathy. The major cause of exudates is leaking of lipids and proteins from damaged retinal blood vessels. Detection of exudates in eye images is used for diagnosis of the maculpathy disease. The proposed framework begins with fuzzy image enhancement of eye images for contrast enhancement in order to better represent objects of the images. Then, the segmentation process is performed to determine the optic disc and blood vessels to remove them. The next step is working on an image with exudates only if existing. A gradient process is performed on the image. The histogram of gradients is evaluated. A cumulative histogram is further generated for discrimination between image with and without exudates. A threshold histogram curve is generated based on predefined images with and without exudates for classification of images in the testing phase. A Convolutional Neural Network (CNN) is used to classify the normal and abnormal cases. The performance of the CNN is higher than traditional networks. The accuracy of the CNN is higher than the accuracy of traditional networks. The main objective of this thesis is to build up an efficient Computer Assisted Diagnosis (CAD) system for the detection of anomalies from medical eye images to help ophthalmologists for identifying diabetic maculpathy, easily. |