The aim of this study was the evaluation of new algorithms to measure diameter of segments of retinal branch vessels offline in local dependence and independent by observer. Methods 360 flashed fundus images (camera FF 450 ZEISS Germany, Visualis IMEDOS GmbH Weimar/Germany) of 12 eyes of healthy volunteers (10 independent sessions containing 3 images for every eye) were analysed. Algorithms detect the vessel diameter along the vessel course automatically. Corresponding segments of a retinal artery and vein were examined (mean length of the segment 2.5 mm) in every image of one eye. Results The marked arterial segment was detected automatically in 359 pictures and the venous segment in all pictures. The mean vessel diameter was detected in single pictures with a mean coefficient of variation (CV) for arteries of 3.4 % and for veins of 2.7 %. The differences of arterial and venous diameter between images were not significant. Analyzing sessions the CV of the mean vessel diameter were reduced for arteries to 2.7 % and for veins to 2.5 %. The standard deviation of the mean vessel diameter was independent of the vessel diameter (branch vessels with diameter of 120 to 200 micrometer). The mean CV of the vessel diameter at single locations were 5.7 % for arteries and 3.8 % for veins. Conclusions The new algorithms are useful for retinal vessel analysis, if there are no questions concerning the dynamic of vessel behaviour.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.