Fractal analysis of normal retinal vascular network.

PURPOSE: In the last decades, various analysis methods with different degrees of effectiveness were proposed in medical imaging. Fractal analysis is a mathematical technique that can quantify the degree of complexity and optimality of vascular networks geometry. The objective of this study is to determine the fractal dimension, applying a new standardized protocol for normal vascular network of the human retina using fractal analysis. MATERIALS AND METHOD: The digitized images of the normal vascular network of the human retina were processed and evaluated with the Image J fractal analysis software and the fractal dimensions of normal retinal vascular network were calculated using the standard box-counting method. RESULTS: The central tendency and dispersion measure of fractal dimension were expressed by the mean value and standard deviation. The mean fractal dimension of normal arteriovenous retinal vascular network was 1.698 +/- 0.003. A comparison with the data from studies performed in the ophthalmologic literature was made. CONCLUSIONS: The retinal vascular fractal dimension is a noninvasive prediction tool that assist researchers and clinicians for early detection of pacients with different retinal vascular diseases.