Analysis of B-mode transverse ultrasound common carotid artery images using contour tracking by particle filtering technique

Abstract

In the recent past, object segmentation plays a vital role in medical image analysis for taking subjective decisions by physicians. Segmentation of the common carotid artery (CCA) images using particle filtering is presented in this paper. Noninvasive B-mode transverse ultra sound images of CCA are used for segmentation. Normally, Ultra sound images are affected by speckle noises. For effective segmentation, the noises are to be removed using preprocessing techniques. An edge preserving anisotropic diffusion filter is used for speckle reduction. In the proposed technique, seed points have been initialized to start the segmentation of the contour. The intensities of the pixels lying inside and outside the contour have been used for obtaining likelihood function of the particle filter. This method is highly effective for images that are affected by speckle noises. The proposed method is suitable for segmentation of CCA wall and diagnosis of atherosclerosis and cardiovascular diseases.