Automatic Segmentation of Liver Tumor in CT Volumes Using Nonlinear Enhancement and Graph Cuts

Aiming at the segmentation challenges caused by low contrast, fuzzy boundary and variant grayscale of liver tumors in abdominal CT images, an automatic liver tumor segmentation method based on nonlinear enhancement and graph cuts is proposed. Firstly, adaptive piecewise nonlinear enhancement and iterative convolution operation are used to improve the contrast of healthy liver parenchyma and tumors according to the gray-level distribution characteristics of liver region. Then, the enhancement result and image edge information are effectively integrated into graph cuts cost computation to segment the liver tumors initially and automatically. Finally, three-dimensional morphological opening operation is performed on the initial segmentation result to remove segmentation errors and increase accuracy. The experimental results on 3Dircadb and XYH databases show that the proposed method can segment liver tumors from abdominal CT volumes effectively and automatically, and the comprehensive segmentation performance of the proposed method is superior to that of several existing methods.