Automated Calculation of Retinal Arteriovenous Ratio for Detection and Monitoring of Cerebrovascular Disease Based on Assessment of Morphological Changes of Retinal Vascular System

In the frame of the ARIC study it could be shown that the retinal vessel system gives important information about retinal, ophthalmic, and cerebrovascular diseases by manually labeling the vessels. In this paper an approach is presented which automatically assesses the retinal vessel systems in fundus images. For this, first the optic disk is located. Afterwards the vessels are segmented and then classified into arteries and veins. In the last step an index, based on the ratio of the diameters of arteries and veins, for the risk of suffering a stroke is calculated. Introduction/Mot ivation According to the WHO stroke is the second frequent cause of death world-wide. In Germany stroke is the third frequent cause of death and the most frequent reason for disability within adults. The treatment of stroke costs 15% of the yearly budget in public health in Germany and there are 180 primary strokes per 100.000 inhabitants each year. So there is a need for a primary stroke prevention in order to decrease the incidence. It is well known that morphological and functional changes in the retinal vessel system are risk indicators for cerebral arteriosclerosis. Using a new quantitative assessment of the retinal vessels a risk-index of stroke, based on morphological parameters, was developed by the ophthalmology group of the Atherosclerosis Risk in Communities Study (ARIC) [I], under the use of conventional ophthalmologic fundus images, which were assessed manually. In our approach a remote risk evaluation for stroke is facilitated by analysing the images of the retinal vessels captured with a non-mydriatic fundus camera automatically. Making this possible, the vessel system of the retina is automatically segmented and classified into arteries and veins. Morphological parameters for the retinal vessels are determined and a risk-index, based on the ratio of arteriolar diameter and venous diameter, is calculated. Using this system it is possible to identify persons with an elevated risk of stroke 'Address: HaberstraUe 2, 91058 Erlangen, Germany. E-mail: