Automated method for tracking vasomotion of intravital microvascular and microlymphatic vessels

Abstract

Dynamic tracking of microvascular and microlymphatic vasomotion presents a critical image processing technique in the evaluation of function and dysfunction of the microvasculature. Many methods for determination of diameter changes have been reported. Previous methods which were specifically developed for vasomotion tracking of intravital, fluorescence-free, rapidly constricting microvascular and microlymphatic vessels have various limitations due to complex image background, vessel wall distortion, image drift, noise and other artifacts. In order to overcome these major obstacles and remove undesirable limitations, this study proposed a tracking strategy based on feature matching of moving object—VasTrack. First, we calculate the image drift vector by feature template matching of a landmark in the background. Second, dynamically regulate the position of a sample line in accordance with the drift vector for sustaining the consistency of measurement location. Third, dynamically recognize and track edge position changes by applying feature template matching using edge context. This method does not require special preprocessing of video image registration and rotation. VasTrack compensates efficiently for image drift and vessel wall distortion, can simultaneously track and determine vessel diameters at any orientation and in multiple locations. Testing proved that VasTrack is robust and accurate, and will satisfy the needs of basic microcirculatory research and clinical inspection.