FULL BODY STATISTICAL SHAPE MODELLING: CURRENT STATE-OF-THE-ART AND CHALLENGES

Statistical Shape Modeling (SSM) is a powerful technique that provides a framework for analyzing and representing the variability of shapes. By capturing the average shape and its natural variations, SSM offers valuable insights into anatomical structures. Statistical shape models have applications in various fields, including medical imaging and computer graphics. A range of methods exists for constructing statistical shape models, from the relatively simple linear Principal Component Analysis (PCA) to more complex approaches like Principal Polynomial Shape Analysis (PPSA) and Gaussian Process Morphable Models (GPMM) [1–3]. Full body SSM specifically aims to capture the variability and anatomical relationships within the entire human body, enabling applications such as personalized medicine, motion analysis, and virtual character creation. Although the idea of possessing statistical shape models of the complete skeleton is appealing, such comprehensive models are currently not readily available. Combining the full skeleton with soft tissue structures, including cartilage, ligaments, muscles, and skin into a comprehensive statistical shape model, has the potential to enhance our understanding of the interdependence of these structures [4–5]. This integrated approach would not only enable to study the functional implications of anatomical variations, but also have applications in personalized surgical planning, human-computer interaction and the creation of more realistic digital characters. We present the challenges and advancements to the current state-of-the-art in the development of an anatomical accurate, comprehensive full-body model, integrating over 180 skeletal components and 1250+ ligamentous constraints. This model can be customized to accurately describe the anatomy of any given patient, thereby incorporating validated anatomical details. The full body anatomical shape model is being constructed based on the public New Mexico Decedent Image Database ( https://nmdid.unm.edu/ ).