Rolling Leukocyte Detection based on Teardrop Shape and the Gradient Inverse Coefficient of Variation

The efficient detection and quantification of rolling leukocytes within intravital microscopy is of both theoretical and practical interest. Currently, methods exist for tracking rolling leukocytes in vivo which rely on manual detection of the cells. In this paper, a new technique inspired by the teardrop shape of the rolling leukocytes is presented for accurately detecting rolling leukocytes. It is based on a feature score, the gradient inverse coefficient of variation (GICOV), which serves to discriminate rolling leukocytes from a cluttered environment. The leukocyte detection process consists of three sequential steps: The first step utilizes a novel teardrop shape matching algorithm to coarsely identify the leukocytes by finding the leukocytes resembling teardrops with a locally maximal GICOV. Second and third steps involve evolution of B-spline snake starting from each of the teardrops found in the first step to refine the leukocytes boundaries and associated GICOV score, and retaining contours with high GICOV scores. The detection process is observed to yield superior results compared to other methods currently in use. The deformation information inherent in the teardrop shapes gives valuable information about the extent to which the cells are adhering to the endothelium.