Effect of different surface processing protocols in three-dimensional images for rapid prototyping

The aim of the study was to evaluate the accuracy of virtual three-dimensional reconstructions of human dry mandibles, produced with different surface processing protocols. Three-dimensional images were built from computed tomography scans of 10 dry mandibles, and the surface was smoothed, refined, or both, generating 30 different images. Linear measurements from anatomical landmarks were calculated and compared with the corresponding measurements of the original dry mandible (gold standard). The results showed no differences between the models that were just refined or just smoothed (p > 0.05), when compared to the gold standard. When these two tools were applied together, there was a statistically significant difference (p < 0.01). In conclusion, the application of a single processing tool (smoothing or refinement) in the virtual models does not affect the anatomical measures. However, the simultaneous application of both tools increases the differences between the reconstructions and the original anatomical parts.