Retinal vascular feature analysis using color fundus imaging

Retinal color imaging has enabled the identification and quantification of retinal vascular features such as vessel caliber, tortuosity and branching angle, etc. Studies have shown that these vascular features have association of pre-clinical or clinical cardiovascular diseases (CVD). In this paper, we discuss the automated measurement of retinal vascular features such as vessel width or caliber, vessel bifurcation and crossover point and vessel tortuosity. Vessel segmentation is the most important step for accurate and efficient vascular feature analysis. Therefore, we have proposed a method for retinal blood vessel segmentation to achieve high accuracy and efficiency. We introduce the existing techniques and their limitations. Following this, we discuss our proposed methods and analysis techniques to extract these vascular features from color retinal images. We also evaluated the results on our proposed feature extraction techniques by comparing other state of art methods.