Description

One of the well-known and commonest diseases that need a computer-aided medical diagnosis is diabetic retinopathy (DR), which leads in most cases to partial or even complete loss of visual capability. The accurate diagnosis of this disease depends upon some features which have to be analyzed in order to quantify the severity level of the disease. Retinal blood vessels are considered as one of the most important features for the detection of DR. The segmentation of blood vessels from the retinal images can be done based on classification or segmentation process. In the segmentation process morphological segmentation and thresholding process were employed. In the classification process based on the features extracted the image pixels were classified into retinal portion or blood vessel portion. The extraction of blood vessels from retinal images is an important and challenging task in medical analysis and diagnosis. Retinal images usually have pathological noise and texture backgrounds, which may cause difficulties in extraction. The line type feature of the blood vessels is not changed when the background textures of the image are different. Identifying and locating blood vessels in fundus images is a critical process since blood vessels were similar in intensity compared to other regions in the images. The process of extraction of the blood vessels is initialized by the preprocessing of the retinal image. Preprocessing is employed in two stages Global Contrast Normalization (GCN) and Zero-phase Component Analysis (ZCA Whitening). The preprocessing step enhances the input images and they are used for the extraction of the features. The preprocessing process is also helpful for the extraction of the feature values from the input images. The blood vessel regions were segmented with the help of Convolutional Neural Network (CNN) classifier. The performance of the process is measured based on performance metrics like Accuracy, Sensitivity, Specificity, Area under Curve and Kappa.