3D Accuracy of Deciduous Tooth Model Reconstructed from CBCT Scan

Abstract

Objective To assess the three dimensional (3D) accuracy of deciduous tooth surface model reconstruction from cone-beam CT(CBCT) scans using image superimposition method. Methods 14 sound extracted human deciduous teeth were scanned by micro-CT and CBCT scanner. The digital tooth models reconstructed from CBCT scans were superimposed onto the standard model from micro-CT scans respectively. Distribution patterns of shape discrepancy were presented using histogram, as well as shown in different colors on the superimposed images. The voluminal discrepancy versus the volume of the standard image (RVD/VS) were calculated and analyzed, using t test. Results There was no statistically significant difference between deciduous anterior teeth and molars(P>0.05). The mean RVD/VS of digital tooth images obtained by CBCT scans was 3.932±4.038%, and significant difference were found (P<0.01), compared with the standard tooth model from micro-CT. Marked by different colors, the superimposed images could give visualized information on the dimensions and distribution patterns of discrepancies. Histogram gave the statistical description about the numbers and magnitude of the discrepancies. Conclusions The superimposition method and the concept of RVD/VS are recommended for 3D accuracy evaluation. CBCT is of high accuracy for tooth modeling and suits for clinical and research purpose.