Abstract

Imaging techniques based on spectral changes of oxygenated hemoglobin (HbO₂) and reduced hemoglobin (Hb) have been used in humans to measure oxygen saturation in retinal vessels. We have been developed a spectroscopic imaging equipment, consists of a normal fundus camera and two CCD cameras. Oxygen saturation imaging of retinal vessels with a clear difference between arteries and veins has been obtained using a morphological processing and a novel data processing of the line convergence index filter. In this study, high resolution CMOS cameras and a non-mydriatic auto fundus camera were applied to achieve accurate extraction of retinal blood vessels and high resolution imaging of oxygen saturation.

© 2011 AOS