Automated System for Retinal Vessel Segmentation

Retinal blood vessels play a significant role in the diagnosis and treatment of various retinal diseases such as diabetic retinopathy, glaucoma, arteriosclerosis and hypertension. Retinal blood vessel extraction is primary step for detecting the above eye diseases. Due to the reason that manual extraction is long and tedious tasks, many automated systems are proposed by various researchers. Automatic segmentation of the retinal vasculature is the major task in the automatic calculation of the features of retinal blood vessels. Screening larger populations for blood vessel abnormalities by eye care specialists can be done using computer based automatic segmentation of blood vessels. In the proposed method we have developed an automated system for the extraction of blood vessels with promising result. In our method we introduced enhanced preprocessing with multi scale line detection for the extraction of blood vessels. The proposed method is evaluated both quantitatively and qualitatively on publicly available DRIVE database. Our proposed method is better than many existing methods by achieving average accuracy, sensitivity and specificity as 0.7421, 0.9803, and 0.9592 respectively.