Estimating patient-specific shape prior for medical image segmentation

Image segmentation is one of the key problems in medical image analysis. This paper presents a new statistical shape model for automatic image segmentation. In contrast to the previous model based segmentation methods, where shape priors are estimated from a general population-based shape model, our proposed method aims to estimate patient-specific shape priors to achieve more accurate segmentation by using manifold learning techniques. The proposed shape prior estimation method is incorporated into a deformable model based framework for image segmentation. The effectiveness of the proposed method has been demonstrated by the experiments on segmenting the prostate from MR images.