An optimized haptic interaction model based on support vector regression for evaluation of endodontic shaping skill

Effective and objective evaluation of endodontic skill is crucial to the interactive simulation of this operation. In this paper, we present a novel evaluation method based on an optimized haptic interaction model characterizing endodontic shaping by applying new statistical learning techniques to this problem. We first present a novel robotic measurement system to collect detailed haptic data during real endodontic shaping operations conducted by experts and establish the needed haptic training set. Then we propose a support vector regression model to estimate the haptic interaction for endodontic shaping. The regression model uses RBF kernel for training, and the optimized parameters of the learned model are obtained by experiments. Applying this model to the virtual endodontic training system, we can evaluate the shaping operations conducted in the virtual environment convincingly.