Detection accuracy of condylar defects in cone beam CT images scanned with different resolutions and units

Abstract

Objectives: To assess the impact of spatial resolution and cone beam CT (CBCT) unit on CBCT images for the detection accuracy of condylar defects.

Methods: 42 temporomandibular joints were scanned, respectively, with the CBCT units ProMax® 3D (Planmeca Oy, Helsinki, Finland) and DCT PRO (Vatech, Co., Ltd., Yongin-Si, Republic of Korea) at normal and high resolutions. Seven dentists evaluated all the test images with respect to the presence or the absence of condylar defects. Receiver operating characteristic curve analysis was employed to define the detection accuracy. Two-way analysis of variance was used to analyse the values under the receiver operating characteristic curves for the differences among imaging groups and observers. Intraobserver variation was analysed using the Wilcoxon test.

Results: Macroscopic anatomy examination revealed that, of the 42 temporomandibular joint condylar surfaces, 18 were normal and 24 had defects on the surface of condyles. No significant differences were found between the images scanned with normal and high resolutions for both CBCT units ProMax 3D (p = 0.119) and DCT PRO (p = 0.740). Significant differences exist between image groups of DCT PRO and ProMax 3D (p < 0.05). Neither the inter- nor the intraobserver variability were significant.

Conclusions: The spatial resolution per se did not have an impact on the detection accuracy of condylar defects. The detection accuracy of condylar defects highly depends on the CBCT unit used for examination.

Keywords: cone-beam computed tomography; mandibular condyle; temporomandibular joint.

Figure 1

(a) An example photograph of condylar defect (solid white arrow). (b) The example of cone beam CT images acquired with ProMax® 3D at normal resolution, (c) ProMax 3D at high resolution, (d) DCT PRO at normal resolution, (e) DCT PRO at high resolution of the same condylar defect. ProMax 3D was obtained from Planmeca Oy, Helsinki, Finland and DCT PRO from Vatech, Co., Ltd., Yongin-Si, Republic of Korea.

Figure 2

Receiver operating characteristic (ROC) curve from the pooled observer performances for each imaging group when the condylar defect was detected. ProMax® 3D was obtained from Planmeca Oy, Helsinki, Finland and DCT PRO from Vatech, Co., Ltd., Yongin-Si, Republic of Korea.