Fractal Matrix and Morphological Analysis of Retinal Vascular Regions

In this scientific paper there is presented a new methodology based on morphological fractal analysis of the retinal vascularisation regions of the eye in order to detect the differences between normal retinal regions and the ones affected by glaucoma and/or diabetic retinopathy. There has been employed differentiation techniques based on fractal dimension computation and network complexity, such as the number of fractal branching points, end points and vessel segments. The necessary information for the statistical evaluations has been extracted from the fractal matrix associated to the analyzed images. The parameters variations in certain retinal regions are correlated with a specific disease, as it can be observed from the results.