07-03-2017, 12:26 PM
Segmentation of the retinal vessel is an essential step in the diagnosis of various eye diseases. An automated tool for the segmentation of blood vessels is useful for eye specialists for the purpose of patient assessment and clinical study. The vascular pattern is usually not visible in the retinal images. In this article we present a method to improve, locate and segment the blood vessels in the images of the retina. We present a method that uses 2-D Gabor wavelet and sharpening filter to improve and sharpen vascular pattern, respectively. This technique locates and segments the blood vessels using algorithms of edge detection and morphological operations. This technique is tested in the publicly available STARE database of manually tagged images that has been established to facilitate comparative studies on the segmentation of blood vessels in retinal images. The validation of our technique of segmentation of retinal imaging vessels is supported by the experimental results.
Automated segmentation of retinal structures allows the ophthalmologist to perform large-scale population monitoring examinations for evaluation of treatment and detection or prevention of retinal diseases. This type of non-intrusive diagnosis present in modern ophthalmology could reduce and prevent blindness and also prevent various cardiovascular diseases worldwide. A retinal blood vessel is the most important and is playing a vital role in the treatment and detection of retinal diseases. The segmentation of the retinal vessel and its delimitation of morphological attributes include width, length, branching or tortuosity pattern and angles for the screening, treatment, diagnosis and estimation of different cardiovascular diseases. Segmenting the blood vessels of the retina using the manual method is one of the tedious and long tasks that also need skills or knowledge and training. It is generally accepted by the medical community that automatic quantification of retinal vessels is the initial step in improving the computer-assisted diagnostic system for ophthalmic disorders. The poor contrast and intensity in the homogeneity of the retinal images affects the marked degradation to the automated segmentation of blood vessel performance methods. The intensity of the retinal images from the background of the image in homogeneity is commonly attributed to the acquisition of images under various lighting conditions.
Automated segmentation of retinal structures allows the ophthalmologist to perform large-scale population monitoring examinations for evaluation of treatment and detection or prevention of retinal diseases. This type of non-intrusive diagnosis present in modern ophthalmology could reduce and prevent blindness and also prevent various cardiovascular diseases worldwide. A retinal blood vessel is the most important and is playing a vital role in the treatment and detection of retinal diseases. The segmentation of the retinal vessel and its delimitation of morphological attributes include width, length, branching or tortuosity pattern and angles for the screening, treatment, diagnosis and estimation of different cardiovascular diseases. Segmenting the blood vessels of the retina using the manual method is one of the tedious and long tasks that also need skills or knowledge and training. It is generally accepted by the medical community that automatic quantification of retinal vessels is the initial step in improving the computer-assisted diagnostic system for ophthalmic disorders. The poor contrast and intensity in the homogeneity of the retinal images affects the marked degradation to the automated segmentation of blood vessel performance methods. The intensity of the retinal images from the background of the image in homogeneity is commonly attributed to the acquisition of images under various lighting conditions.