Statistical Surface Wavelets Model for Multiscal and Spatial-Localized Medical Image Analysis

Statistical shape models which represent the shape variations within a population are used in a variety of medical image applications, such as statistical shape analysis, model-guided segmentation and registration. This paper presents our statistical surface wavelets model that has three highly desirable properties of a shape model: compact representation, multi-scale shape description, and spatial-localization of the shape variation. To the best of our knowledge, there is no known prior work that achieves all these properties simultaneously. We have implemented our model to perform shape analysis. Preliminary results achieved from the cerebral lateral ventricle and caudate nucleus are encouraging.