Automatic processing and analysis of video data formed by a capillaroscope

Abstract

The problem of the processing and automatic analysis of video information formed by a computer capillaroscope has been investigated. The specific features of original video data have been studied, an algorithm to eliminate the drift of frames and form an averaged image has been proposed. The problem of detecting and analyzing the capillaries has been solved, which is comprised of the stages of forming and filtering the map of contour lines of capillaries, the syntactic analysis of contours, the selection of the major capillary, and the prepartion and analysis of the morphological characteristics of contour lines. A homeomorphic straightening mapping of the capillary area into rectangular shape area is proposed. The transformed data are used to determine the characteristics of capillary blood flow. An algorithm for measuring the velocity of instantaneous blood flow as a function of time and location along the capillary has been developed.