Validation of a Computer Aided Segmentation System for Retinography

Abstract

In spite of the huge literature on angiography, some problems are still open to discussion, such as segmentation of entire vascular networks. In  the present work a new computer approach is developed in two stages, with the aim of improving the analysis and comparison of retina vessel images in the follow up of patients. The first stage adopts multiscale filtering to detect objects of different sizes: a two scale Laplacian of Gaussian scheme is used with the related sigma values chosen according to the smallest and greatest vessel widths. An approximate segmentation is achieved simply by means of the Laplacian sign. The interpretation stage is application-specific and accomplishes classification and quantitative analysis. The skeleton of the binary structures is subdivided in vessel segments, their features (intensity, position, length and width) are fed into an artificial neural network (ANN), after back-propagation training. The segments classified as vessels are assembled into the retinal vascular tree by rule-based tracking, starting from optic disc (OD). Results are evaluated on STARE and DRIVE data bases. Accuracy is 95% and the false positive rate is decreased to about 1%, lower than literature values.