A Fast Hybrid Method for Interactive Liver Segmentation

Accurate liver segmentation from abdominal computed tomography (CT) images is one of the most important steps for computer aided diagnosis (CAD) for liver CT. Recently, interactive segmentation plays an important role in liver segmentation. In this paper we propose a fast hybrid method for liver segmentation from abdominal CT image. Firstly, the CT image is enhanced and denoised by linear stretch and anisotropic diffusion. Secondly, in order to reduce the computation cost, watershed transform is used for partitioning the image into small region pieces. Thirdly, the image region graph is constructed based on the watershed pre-segment region using typical Gaussian weighting energy function. At last, random walk algorithm is applied to obtain the final segmentation results. The experiments on 2D CT images show that the proposed method achieves high segmentation accuracy and runs quite fast.