AN AUTOMATIC SEGMENTATION METHOD APPLIED TO IMAGES OF IV-OCT IN HUMAN ARTERIES WITH ATHEROSCLEROTIC PLAQUES

Intravascular optical coherence tomography (IV-OCT) is an imaging modality in-vivo based on the introduction of an intravascular catheter for viewing the inner wall of blood vessels with a spatial resolution of 10μm to 20μm. The acquisition and analysis of vessel images by IV-OCT are recent and enables several new applications and the development of automatic techniques of computer vision and related areas. This paper aims to describe a fully-automatic method to compute vessel's lumen area in IV-OCT images with challenges including the presence of catheters, guide wires and atherosclerotic plaques. Tests were performed in 787 human IV-OCT images presenting a mean difference in area of 0.04 mm 2 comparing with manual segmentations. The results raise the possibility that automatic methods have to perform a subsequent lumen