Adaptive Level Set Method for Segmentation of Liver Tumors in Minimally Invasive Surgery Using Ultrasound Images

Ultrasound images have been employed in guiding clinical interventional therapy procedures for liver tumor. However, segmenting liver tumor in the ultrasound images presents a unique challenge because of the low-contrast objects in the noisy image. Snakes, or active contours have had limited success in such noisy and complex image. In this paper, an adaptive level set method is proposed, which combines the global statistics and boundary statistics instead of image gradient and edge strength .Compared to traditional level set method, the experiment results show that the proposed level set method was feasible , enabled accurate and robust.