Abstract
Statement of problem. The accuracy of the external surface and internal trabecular architecture of large cone beam computed tomography (CBCT)-derived dentomaxillofacial anatomic casts has not yet been thoroughly investigated.
Purpose. The purpose of this comparative study was to evaluate the quantitative accuracy of CBCTderived mandibular casts by applying an innovative land-mark free methodology.
Material and methods. Following inclusion and exclusion criteria, a CBCT scan of an 18-year-old woman was acquired. The mandible was segmented and isolated from the data set. The segmented mandible included depiction of the cortical surface, trabecular architecture, erupted teeth, and impacted third molars with incomplete root formation. Fifteen mandibular casts were fabricated by using multijet (MJ=4), digital light processing (DLP=4), stereolithography (SLA=2), fused deposition modeling (FDM=2), colorjet (CJ=2), and selective laser sintering (LS=1)-based high-quality medical commercial and office printers. Each printed cast was scanned and superimposed onto the original mandible, and the accuracy of the complete mandible and individual surfaces were assessed with a color-coded map.
Results. When the overall combined error associated with complete casts based on printing technology were compared, MJ showed the highest accuracy (0.6 +/- 0.7 mm). FDM technology (2.2 +/- 3.4 mm) had the highest overall absolute mean difference. No significant difference was observed when both individual surfaces and the complete mandible were compared.
Conclusions. Overall, casts replicated the skeletal and dental anatomic surfaces well. However, shortcomings were observed in relation to depicting trabecular architecture.