Automatic Extraction of Lesions and Hepatic Structures in Liver using Segmentation Techniques

摘要

Computed Tomography is an important instrument that can be used to analyze the images of the human organs such as liver. It is helpful to diagnose certain liver diseases and applicable for surgical planning. In some cases, it seems to be difficult to obtain the accurate shape and intensity of the liver. In this paper, a hybrid approach is proposed by combining the seed point selection method and connected component method with watershed algorithm to obtain a better accuracy and robustness of the CT images. The experimental result shows that the proposed hybrid approach is fast and efficient in detection with a high accuracy of 98.6 %.