A Low-Interaction Automatic 3D Liver Segmentation Method Using Computed Tomography for Selective Internal Radiation Therapy
Figure 2
Importance of selection of seeds points for the k-means algorithm; (a) 5 masks M1–M5 obtained using random selected seed points, (b) 5 masks M1–M5 obtained using uniformly selected seed points, and (c) 5 masks M1–M5 obtained using user selected seed points. Results show that a user identification of the 5 regions as outlined in this study results in a better demarcation of liver tissue.