الفهرس | Only 14 pages are availabe for public view |
Abstract Medical imaging is currently undergoing a rapid development with a strong emphasis being placed on the use of the imaging technologies to render surgical and therapeutic procedures to improve the accuracy. Retinal imaging is the technique of creating visual representations of the interior of the eye. It has a vital part of ophthalmology practice. It helps in the early detection of vision disorders and diseases that can affect the eye such as diabetes, hypertension, glaucoma, and macular degeneration. Diabetic retinopathy (DR) is a condition, where the retina is damaged due to fluid leaking from the blood vessels into the retina. In extreme cases, the patient will become blind. Therefore, early detection of diabetic retinopathy is crucial to prevent blindness. Although retinal imaging is a common clinical procedure used to determine if a patient suffers from DR, there are several artifacts that affect retinal images and cause poor quality of them. Images with camera artifacts can lead to false diagnostics. Artifacts are caused by poor illumination, degradations coming from blurring. The main objective of this thesis is to overcome artifacts in fundus images using different denoising techniques and implement automated detection algorithms to process a large number of fundus images captured from mass screening of diabetic patients with high accuracy. The automated detection system is used to detect the retinal features such as blood vessels area, microaneurysms area, exudates area and texture features. These features are fed to an artificial neural network (ANN) classifier. The ANN classifies the data into two categories; normal and abnormal. A proposed hybrid detection algorithm for automatic feature detection based on entropy and homogeneity is used to improve the performance of the automated system and increase the specificity and accuracy. The proposed system is considered as an automatic tool that can aid ophthalmologists to diagnose and screen diabetic retinopathy. |