Generation of Point-Based 3D Statistical Shape Models for Anatomical Objects

A novel method that allows the development of surface point based three-dimensional statistical shape models is presented. Given a set of medical objects, a statistical shape model can be obtained by Principal Component Analysis. This technique requires that a set of complex shaped objects is represented as a set of vectors that uniquely determine the shapes of the objects, and at the same time are suitable for a statistical analysis. The correspondence between the vector components and the respective shape features has to be identical in order for all shape parameter vectors to be considered. We present a novel approach to the correspondence problem for arbitrary three-dimensional objects, which involves developing a template shape and fitting this template to all objects to be analyzed. The method is successfully applied to obtain a statistical shape model for the lumbar vertebrae.