[Preoperative accuracy of selective laser sintering (SLS) in craniofacial 3D modeling: comparison with patient CT data]

Abstract

Purpose: To evaluate the accuracy of craniofacial 3D models produced in the standardized selective laser sintering (SLS) technique from multislice computed tomography (MSCT) data sets in comparison with patient data and to investigate the effect of potential causes of inaccuracies.

Materials and methods: 19 models were considered and examined by MSCT. The patient CT data used for 3D modeling was analyzed and compared to the 3D model data. 15 anatomical landmarks were defined and 20 distances were digitally measured. The digital measurements of both CT data sets were compared to manually measured distances of the SLS model.

Results: There was not a statistically significant difference (p < 0.05) between the measurements of the distances concerning all three groups (patient CT data, model CT data, manual measurement of the model). The mean values of the differences were between 0.5455 and -0.3214 mm.

Conclusion: We found a high accuracy of SLS 3D models, which is due to a high precision in the modeling process and to the small voxel size of patient CT data achieved by MSCT. Anatomical landmarks in patient and model CT data sets and on the 3D model were able to be accurately reproduced, which is important for preoperative planning.