China Journal of Oral and Maxillofacial Surgery ›› 2019, Vol. 17 ›› Issue (3): 221-224.doi: 10.19438/j.cjoms.2019.03.006

• Original Articles • Previous Articles     Next Articles

Accuracy of digital models reconstructed by cone-beam CT under different segmentation thresholds

SHEN Yu-qing1, LIU Wen-juan2, SUN Jian1, YU You-cheng1, ZHOU Qian-rong1, XU Yi1   

  1. 1.Department of Stomatology, Zhongshan Hospital Fudan University. Shanghai, 200032;
    2.Department of Stomatology, Xuhui Central Hospital. Shanghai. 200031, China
  • Received:2019-03-20 Revised:2019-04-03 Online:2019-05-20 Published:2019-06-21

Abstract: PURPOSE: To compare the accuracy of digital models reconstructed by cone-beam CT(CBCT) under different segmentation thresholds (STs). METHODS: A total of 20 patients who had CBCT images for diagnostic purposes were included in this study. Three-dimensional digital models were reconstructed by CBCT scans using different STs(based on gray value (GV) from 200 to 1800,gap 200)with Mimics 15.01. Plaster models were obtained, and mesiodistal diameter of each tooth was measured in all models. Paired t test was used to evaluate the accuracy of the digital models in comparison to the plaster models. Scatter plots and intraclass correlation coefficient (ICC) were also used to evaluate the correlation between the models with SPSS 16.0 software package. RESULTS: The mean difference between digital models and plaster model was from-0.36 mm to 0.32 mm. There was no significant difference between the plaster model and the digital models reconstructed under ST GV 1200 and 1400 (P>0.01). The measurements of digital models reconstructed under other STs were different from the plaster ones (P<0.01). ICCs between digital models and plaster model were all >0.95. CONCLUSIONS: The accuracy of CBCT reconstructed 3-dimensional digital model is influenced by different STs. Therefore, selection needs to be made in clinical application.

Key words: Cone-beam CT, Digital model, Segmentation threshold, Gray value, Accuracy

CLC Number: