Fractal analysis of normal retinal vascular network

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.

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.

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.

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.