A Hybrid Approach for Detection of Tumor in Abdominal CT Scan Images
In this paper an efficient segmentation technique for detection of tumor in abdomen/liver CT scan images is
presented. The proposed method integrates region growing and level set techniques. First, the contrast of original image is
enhanced making the boundaries of the object apparent for segmentation; secondly, region growing combined with anatomical
prior knowledge is applied to extract abdomen/liver region; thirdly, levelset method is used for refinement; and finally,
morphological operations are used for post-processing. Standard deviation and tumor burden parameters are used to validate
the proposed method. The experimental results show that the proposed method achieves an accuracy of 91.66% and specificity
Index Terms - Abdomen/Liver Segmentation, Region Growing, Levelset.