Can an entry-level 3D printer create high-quality anatomical models? Accuracy assessment of mandibular models printed...

This study was performed to determine whether an in-house printed mandible model is sufficiently accurate for daily clinical practice. Ten example mandible models were produced with a desktop 3D printer (fused filament fabrication, FFF) and compared with 10 equivalent mandible models fabricated using a professional-grade 3D printer (selective laser sintering, SLS). To determine the precision of the printed models, each model was scanned with an optical scanner. Subsequently, every model was compared to its original standard tessellation language (STL) file and to its corresponding analogue. Mean±standard deviation and median (interquartile range) differences were calculated. Overall these were -0.019±0.219mm and -0.007 (-0.129 to 0.107) mm for all 10 pairs. Furthermore, correlation of all printed models to their original STL files showed a high level of accuracy. Comparison of the SLS models with their STL files revealed a mean difference of -0.036±0.114mm and median difference of -0.028 (-0.093 to 0.030) mm. Comparison of the FFF models with their STL files yielded a mean difference of -0.055±0.227mm and median difference of -0.022 (-0.153 to 0.065) mm. The study findings confirm that in-house 3D printed mandible models are economically favourable as well as suitable substitutes for professional-grade models, in particular considering the geometric aspects.