An Automatic Approach For Segmentation of Ultrasound Liver Images
Pradeep Kumar · 2013
The liver is a large, pyramidal shaped organ and defined using real time ultrasound imaging. Normal ultrasound images provide a good insight of the disease. But there are a few concerns like the expertise needed and clear distinction of diseases. In the present study full automatic segmentation has been proposed and using statistical features helps to distinguish between normal and ultrasonic tumor liver images. The automatic segmentation of anatomical structures in ultrasound imagery is a real challenge due to speckle noise and artifacts which are inherent in these images. The noise is filtered using peak and valley method, image is smoothened and again a second stage filter is adopted to further filter the noise and to improve the quality of image.