Cone detection and blood vessel segmentation on AO retinal images

With the advent of Adaptive Optics (AO) for the high-resolution imaging of the retina, it is possible to study the individual photoreceptors and their spatial distribution in vivo. The presence of the blood vessels affects negatively the detections, and for this reason clinicians select manually the regions devoid from the vessels to be analysed. We present here a method that we developed for the segmentation of blood vessels in AO retinal images. With the choice of a suitable cone detection algorithm, we are now able to automatically analyse the images acquired by an AO fundus camera in their entirety.