A whole body statistical shape model for radio frequency simulation

The development of ultra low power wireless sensors for customized wearable and implantable medical devices requires patient specific models for radio frequency simulation to understand wave propagation in the body. In practice, the creation of a patient specific whole-body model is difficult and time consuming to create. It is therefore necessary to establish a method for studying a population in a statistical manner. In this paper, we present a statistical shape model for the whole body for RF simulation. It is built from 10 male and 10 female subjects of varying size and height. This model has the ability to instantiate a new surface mesh with the parameters allowed by the training set. This model would provide shapes of varying sizes for studies, without the requirement of obtaining subject specific whole body models. Results from finite-differences time-domain simulation are presented on the extreme shapes from the model and demonstrate the need for a full understanding of the range in body shapes.