The increasing importance of three-dimensional imaging in medicine leads to a growing demand for volumetric image analysis and automatic segmentation. Due to their robust performance, statistical shape models trained on a collection of example data are especially suited for that purpose. In this book, a three-step procedure for generating these models and employing them for 3D segmentation is presented. The first step is the identification of corresponding landmarks on the example data, required for training the geometric models. The second step consists of modeling the appearance, i.e. gray-value environment, of the object of interest. The final step integrates shape and appearance model into a robust search algorithm to analyze new images. The presented methods were evaluated on three medical applications: segmentation of the liver in CT data, of the lung in MRI data, and of the prostate in ultrasound images. This book is targeted towards graduate students and researchers in biomedical image analysis who want to gain in-depth insight into the field of statistical shape modeling.