Abstract: Understanding the functionality of microcirculation is a key factor in the analysis of blood circulatory system. The blood flow distribution changes, based on the physiological effects of disorders. This study presents a method for analysis of microcirculation images captured from lingual surface of 10 animal subjects. The technique applies advanced image processing methods to stabilize images, segment micro vessels (capillaries and small blood vessels), and estimate the average Functional Capillary Density on 20 consecutive frames for each subject. The algorithm consists of four main parts: preprocessing, image stabilization, entropic-based adaptive local thresholding segmentation and post-processing. The key objective is to quantitatively examine the changes that occur in microcirculation over treatment periods for diseases as well as for the resuscitation process. The designed system will help physicians and medical researchers in diagnostic and therapeutic decision making to determine the sufficiency of resuscitation process and the effect of drug consumption in patients. In particular, the system focuses on minimizing user interaction while improving the accuracy of the analysis. Visual evaluation of the results by medical experts indicates that the technique is capable of identifying 95% of active capillaries and blood vessels in images.

Keywords: Diagnostic Analysis, Microcirculatory Images, Blood Vessels, Entropic Thresholding.