Ultrasound image segmentation for liver tumor

Abstract

The purpose of this study to develop and compare between processing issues when dealing with segmentation of ultrasound images. Two main techniques were examined and results were shown for both Level Set segmentation and Snake algorithm. Ultrasound images of liver tumor were taken into account, for each tissue has its own individual characterization. Speckle noise and tumor were to be differentiated. Error computation and graphics have helped in the discussion of distinction between different parameters and approaches to end up with the most accurate means when dealing with imaging.