Amr R. Abdel-Dayem and Mahmoud R. El-Sakka, "Diffusion-based Detection of Carotid Artery Lumen from Ultrasound Images", International Conference on Image Analysis and Recognition, ICIAR'2009, LNCS 5627, pp. 782-791, Springer-Verlag Berlin Heidelberg, July 2009, Halifax, Canada.

Abstract

This paper presents an experimental study on the effect of using diffusion-based filters on segmenting carotid artery ultrasound images. Moreover, comparisons with other segmentation schemes, found in literature, were conducted. In this study, the segmentation process starts with the original ultrasound image as the initial image uo (the image at time t=0). Then, the image diffuses as the time t advances until a steady state is reached. At steady state, the real component of the diffused image will be a smoothed version of the input image, whereas the imaginary component will approximate a smoothed second derivative, which is used to extract the artery contours. The experimental results demonstrated the efficiency of diffusion-based filters in segmenting carotid artery ultrasound images.