The paper presents the automated region based segmentation of liver tumor and hepatic vein from abdominal CTA images of diseased populations with inconsistent enhancement. In this paper, the liver is first segmented from abdomen CTA images using fast and reliable semi-automatic method localizing region-based active contours. A new fuzzy Bernstein polynomial level set algorithm is proposed for fast and efficient segmentation of liver lesions and hepatic vein. The region of interest is represented in a lower dimensional subspace using a set of pre-defined Bernstein basis functions. The initial level set is able to evolve directly from the initial segmentation by spatial fuzzy clustering to segment lesions and hepatic vein. The performance evaluation of the proposed algorithm was carried out on the dataset of 10 liver tumor patients. The performance comparison of the proposed algorithm with other state of art algorithms demonstrates its robustness in segmenting low contrast and inhomogeneous CTA images.