Liver Segmentation from CT Images Based on Region Growing Method
Yufei Chen, Zhicheng Wang, Weidong Zhao, Xiaochun Yang · 2009
Accurate liver segmentation on computed tomography (CT) images is a challenging task because of inter and intra- patient variations in liver shapes, similar intensity with its nearby organs. We proposed a liver segmentation method based on region growing approach. First of all, basic theory of region growing approach is introduced. Secondly, a pre-processing method using anisotropic filter and Gaussian function is employed to form liver likelihood images for the following segmentation. Thirdly, an improved slice-to-slice region growing method combined with centroid detection and intensity distribution analysis is proposed. Finally, the superior liver region is extracted by applying the morphologic operation. Experiments on a variety of CT images show the effectiveness and efficiency of the proposed method.