Extraction of Retinal Blood Vessels from Retinal Fundus Image for Computer Aided Diagnosis

Changes in the retinal blood vessel features are the first indications of many cardio-vascular disease. Detection and analysis of these changes can help the patient to take the early action to arrest the progression of the disease. Automation of this process will not only reduce the time required but also decrease the cost associated with trained clinicians and remove the factor of human error associated with manual grading. In this paper we propose a simple, fast and efficient retinal blood vessel algorithm which uses an image processing concept called as mathematical morphology. The developed algorithm was tested on two publicly available databases DRIVE and STARE. The results were then quantitatively evaluated in terms of values of sensitivity, specificity and accuracy. High values of evaluation parameters and negligible computation proved the feasibility of the application of the algorithm in real time. The main advantage of the algorithm is that it performs well at extracting vessels even from pathological images