Anotace:
Remodeling of a microvascular network is common part of pathological changes associated with wide spectrum of diseases. Quantitative analysis of these alterations relies often on analysis of a point-pattern on the histological slide, i.e. on sections through the microvascular network only. Common techniques are based on the estimation of the average density of points representing section through microvessels on the histological image. This approach inherently omits the information about the regularity of the pattern. Thus, we used approach based on the Voronoi segmentation and chose the best statistical model of areas of Voronoi cells surrounding microvessels on 20 samples of human myocardium. The best model is based on the log-normal distribution. Parameters of the model for given data can be estimated as a mean and a standard deviation of logarithms of areas of Voronoi cells. Moreover, these parameters can be transformed to the widely used measure called the microvascular density.