Width Measurement from True Retinal Blood Vessels

A retinal image can establish the state of human body by simply providing the information about what is happening inside it. Because the AVR status of blood vessels is having good correlation with hypertension, coronary, heart disease and stroke. However they require the accurate extraction of distinct blood vessels and width measurement. The Existing width measurement techniques met with a challenging problem due to ambiguities caused by bifurcations, crossovers and the mis-identification of vessel caliber. In this paper, Graph Tracer Algorithm is introduced to deal with above measurement conflicts which means, it identify true blood vessels and its appropriate bifurcations and cross overs. This algorithm is a post processing step which is followed by the segmentation technique, Multi-Scale Line Tracking (MSLT). Finally Identified blood vessels are subjected into the width measurement with the help of vessel caliber annotation tool. The proposed measurement technique is giving 94.6% of accuracy and this method provides extra 15% of accuracy than the normal measuring technique. Keyword – hypertension, bifurcations, cardiovascular diseases, clinical diagnosis, segmentation technique, annotation tool.