Generating Classes of 3D Virtual Mandibles for AR-Based Medical Simulation

Abstract

Simulation and modeling represent promising tools for several application domains from engineering to forensic science and medicine. Advances in 3D imaging technology convey paradigms such as augmented reality (AR) and mixed reality inside promising simulation tools for the training industry. Motivated by the requirement for superimposing anatomically correct 3D models on a Human Patient Simulator (HPS) and visualizing them in an AR environment, the purpose of this research effort is to derive method for scaling a source human mandible to a target human mandible. Results show that, given a distance between two same landmarks on two different mandibles, a relative scaling factor may be computed. Using this scaling factor, results show that a 3D virtual mandible model can be made morphometrically equivalent to a real target-specific mandible within a 1.30 millimeter average error bound. The virtual mandible may be further used as a reference target for registering other anatomical models, such as the lungs, on the HPS. Such registration will be made possible by physical constraints among the mandible and the spinal column in the horizontal normal rest position.