Segmentation of Region of Interest in Ultrasound Images Using Fuzzy Logic Clustering Method

Abstract

Ultrasound image analysis is desirable for diagnosis of presence stone, cyst and other abnormalities, whereas the presence of speckle noise in ultrasound images may degrade visual quality and also hinder the diagnosis procedure. The present study is aimed to detect stone region in the image. It is important to suppression speckle noise to enhance the visual quality. Moreover, noise suppression is prerequisite condition for object segmentation in ultrasound images. To find the region of interest, it is important to segment the object from the background.  The fuzzy logic clustering method has been used to represent region of interest and background. It is flexible clustering method which allows the elements in different clusters to be members of different groups simultaneously. It helps to justify the members of group which lie on the edge of two groups. As stone region has the similar feature characteristic as of non-stone region due to presence of speck noise everywhere. Such case fuzzy clustering method has shown better results.

Keywords:Kidney stone detection, segmentation, region of interest detection, fuzzy logic method