Estimating the internal elastic membrane cross-sectional area of coronary arteries autonomously using optical coherence tomography images

Abstract

We describe a method to autonomously identify segments of the internal elastic membrane (IEM) in optical coherence tomography (OCT) images and estimate the area bounded by the IEM, the IEM cross-sectional area (CSA). The method relies upon a process by which the lumen and appropriate region of interest are determined, standard filtering and edge-detection algorithms are applied to the image, extraneous identified edges are eliminated based upon edge and adjacent image characteristics, and an ellipse is fit to the surviving IEM edge candidates. The method as implemented resulted in CSA estimates that differed by 3.1±1.3 mm2-38% smaller, on average-relative to gold standard measurements made by an experienced expert interventional cardiologist in intravascular ultrasound (IVUS) images.