Retinal vessel segmentation in fluorescein angiograms using a commitee of local experts

It is well-known that Age Related Macular Degeneration is the main cause of visual impairment in industrialized countries, while demonstrating an increasing trend. Fluorescein Angiography is the preferred technique used in the diagnosis, prognosis and in following up the development of the disease. A key task, in order to perform automated analysis of angiograms, is segmentation of the vascular network. This is particularly challenging in fluorescein angiograms due to the characteristics of the angiogram, namely, uneven illumination, large intensity variations within vessels, low contrast between the background and the vessels, and unpredictable blood filling patterns. This work presents a methodology for segmentation of the vascular network in fluorescein angiograms, based on the analysis of the first- and second-order directional derivatives and phase congruency information. This is used by a committee of local experts, where each expert is represented by an artificial neural network. A fuzzy clustering algorithm is used for expert selection based on the specific input pattern. The output of the system is determined through a winner-takes-all rule. The performance of this approach is demonstrated on a dataset, which contains images of both normal fundus and fluorescein angiograms.