[Detection accuracy of occlusal caries by cone-beam computed tomography images scanned with different parameters].

OBJECTIVE

To compare the detection accuracy of occlusal caries in cone-beam computed tomography (CBCT) images scanned with different scanning parameters.

METHODS

Thirty-seven extracted human premolars and molars without restorations or obvious cavities were selected for this study. The teeth 3 or 4 as a group were mounted in 12 plaster blocks and scanned with DCT Pro (VATECH, Co., Ltd., Yongin-Si, S.Korea) at normal and high resolution settings and with ProMax 3D (Planmeca Oy, Helsinki, Finland) at low, normal and high resolution settings. In addition, the tooth blocks were imaged with the ProMax 3D at three different tube currents. Ten doctoral candidates of Peking University School and Hospital of Stomatology evaluated all the images of the tooth blocks using a five-level confident scale. Actual presence and the extent of caries were established by histological examinations. The areas under the ROC curves (Az value) were analyzed.

RESULTS

Eleven (29.7%, 11/37) teeth were sound, 13 (35.1%, 13/37) had enamel caries and 13 (35.1%, 13/37) had dentine caries. There were no significant differences between normal resolution and high resolution for DCT Pro on the detection accuracy of occlusal (enamel+dentine) caries (0.698 ± 0.064 vs. 0.735 ± 0.044, P>0.05).No significant differences were found for ProMax 3D among low, normal and high resolution (0.700 ± 0.031 vs. 0.700±0.054 vs. 0.701 ± 0.041, P>0.05). For dentinal caries, there were no significant differences between CBCT images scanned with different resolutions for DCT Pro(0.776 ± 0.078 vs. 0.811 ± 0.047, P>0.05) or ProMax 3D (0.713 ± 0.039 vs. 0.685 ± 0.063 vs. 0.713 ± 0.040, P>0.05). No significant differences were found for enamel caries detection for DCT Pro (0.620 ± 0.068 vs. 0.659 ± 0.048, P>0.05)or ProMax 3D (0.686 ± 0.050 vs. 0.685 ± 0.063 vs. 0.689 ± 0.063, P>0.05). For ProMax 3D, there were no significant differences among different tube currents on the detection accuracy of occlusal caries (0.653 ± 0.065 vs. 0.700 ± 0.054 vs. 0.67 ± 0.062, P>0.05).

CONCLUSION

Different resolutions did not have an impact on the detection accuracy of occlusal caries. The tube currents for ProMax 3D did not show any effect on occlusal caries detection.