Liver Segmentation with Semi-Supervised Learning

Efficient liver segmentation from volume data provides important assistance for minimal invasive surgery and treatment. However, this task suffers from the special anatomy and topological changes. This paper presents a robust interactive method, which treats it as a semi-supervised learning task. An initial classification is performed to partition the volume data into homogeneous blocks to guide the segmentation. It is easy to implement and a more general linear or nonlinear model can be formed by virtue of semi-supervised learning. Experimental results demonstrate the performance of the proposed scheme in liver contours extracting.