AUTOMATIC SEGMENTATION OF INTRAVASCULAR ULTRASOUND IMAGES - A TEXTURE-BASED APPROACH

Extraction of blood vessel boundaries from intravascular ultrasound im ages is essential in the quantitative analysis of cardiovascular funct ions. In this study, we are presenting a completely automated procedur e for determining blood Vessel borders. This approach uses textural op erators to separate different tissue regions and morphological process ing to refine extracted contours. The method was tested in a set of 29 intravascular ultrasound images obtained in vivo. To assess the perfo rmance of the method, we have compared the automatically processed ima ges with the manual tracings, using three different criteria: correlat ion coefficient, match ratio, and relative error of computed shape par ameters. In both contour detection and shape parameters estimation, th e proposed method yielded consistently good results. Due to its robust ness and accuracy, this approach is appropriate for clinical use, wher eas computational efficiency of the method facilitates low-cost implem entation.