Analysis of intima media thickness in ultrasound carotid artery images using level set segmentation without re-initialization

Carotid wall analysis using ultrasound image is a challenging task for the diagnosis of cardiovascular pathologies. Intima-Media Thickness (IMT) is the marker of the progression of atheroscelerosis in carotid arteries which leads to stroke. In this paper, an attempt has been made to detect the IMT of far wall, using level set segmentation method based on edge map without re-initialisation. Geometric features such as equivalent diameter, solidity and extent are extracted from the normal and abnormal images. Results show that the proposed method is able to extract the boundary of far wall despite the presence of noise in ultrasound images. The features show distinct variation between normal and abnormal images. Its values are higher in abnormal compared to normal images. Among all the features, extent shows high variation between normal and abnormal images. Hence, this method of analysis could be used in early identification of cardiovascular diseases.