STOCHASTIC PCA-BASED BONE MODELS FROM INVERSE TRANSFORM SAMPLING
When developed on bones, Statistical Shape Models (SSM) and Statistical Intensity Models (SIM) respectively describe the mean shape and mean density distribution identified within a certain population as well as the main modes of variations of shape and density distribution from their mean values, i.e. the shape and density main features. The availability of this kind of models, carrying information about the detailed anatomy and structure of bones, provides considerable new opportunities for diagnosis, evaluation, treatment of skeletal disorders, as well as for the design patient-specific devices, among others. The potential of SSM and SIM has recently been recognized within the bone research community, and a number of studies on the topic have indeed been published.
Focusing on SSM, the possibility to have three-dimensional geometries of a bone district often depends on the availability of CT images. In this context, the lack of clinical images could therefore be overcome taking advantage of a previously built SSM, which potentially allows the extraction of an infinite number of different shapes.
Hence, herein we make available for the whole scientific community the statistical models of the mandible and of the proximal femur, able to generate geometrical models upon request.
Stochastic Mandible Generator 
(2.36 MB)
Stochastic Proximal Femur Generator (1.88 MB)
Please, if using the here proposed SSM for geometric models generation, don't forget to give credits to this work, citing the following reference:
G. Pascoletti, A. Aldieri, M. Terzini, P. Bhattacharya, M. Calì, E.M. Zanetti. Stochastic PCA-based bone models from inverse transform sampling: proof of concept for mandibles and proximal femurs. Applied Sciences
People
