Unified skinning of rigid and deformable models for anatomical simulations

We propose a novel geometric skinning approach that unifies geometric blending for rigid-body models with embedded surfaces for finite-element models. The resulting skinning method provides flexibility for modelers and animators to select the desired dynamic degrees-of-freedom through a combination of coupled rigid and deformable structures connected to a single skin mesh that is influenced by all dynamic components. The approach is particularly useful for anatomical models that include a mix of hard structures (bones) and soft tissues (muscles, tendons). We demonstrate our skinning method for an upper airway model and create first-of-its-kind simulations of swallowing and speech acoustics that are generated by muscle-driven biomechanical models of the oral anatomy.