Automatically generating subject-specific functional tooth surfaces using virtual mastication.

High-accuracy geometrical models of a subject's mandibular and maxillary teeth are combined with recordings of natural chewing trajectories of the same subject to obtain a subject-specific virtual model of mastication-the virtual masticator. The virtual masticator and a shape-optimization algorithm, which is based on removing collisions occurring between a generic maxillary tooth/teeth and the mandibular antagonists during mastication, is used to automatically reconstruct functional tooth surfaces. The process was tested using a chewing trajectory stemming from recordings made of an individual while eating elastic-type foods, a generic maxillary tooth, and the mandibular second molar of that individual. Comparing the obtained results with the actual tooth, the errors within the occlusal and functional regions of the the right second maxillary molar range between -90 and 200 mum and these errors do not change any more after three chewing cycles. These results indicate that a small number of chewing cycles is sufficient to remove occlusal interferences in the virtual tooth model. Such automatically reconstructed tooth surfaces can provide guidance during the design stage of dental fixed restorations manufactured using computer-aided design and manufacturing (CAD/CAM) systems and provide additional information for the design of dental implants or bridges.