Effect of Slice Thickness on the Accuracy of Linear Measurements Made on Cone Beam Computed Tomography Images (InVitro)

Abstract

Objective: Cone beam computed tomography (CBCT) is applied for the imaging of the maxillofacial and dental structures, particularly for surgical treatments and dental implants. The aim of this study was to assess the effect of slice thickness on the accuracy of CBCT linear measurements. Methods: In this diagnostic accuracy study, forty-two titanium pins with the same dimensions were inserted into seven dry sheep mandibles. The length of the pins before the insertion was carefully measured by a digital caliper, (0.01mm accuracy). Imaging of the mandible performed using CBCT New Tom VGi. After image reconstruction by NNT Viewer, linear measurements were made on cross-sectional slices (thicknesses of 0.125, 0.5, 1 and 2mm) by three radiologists. The accuracy of measurements assessed using descriptive indices and compared between different slice thicknesses by repeated measures ANOVA.Results: Repeated measures ANOVA showed a significant difference between different slice thicknesses (P=0.024). According to the least significant difference (LSD) test, the difference in absolute errors was significant in all thicknesses (P=0.024). Measurements at 0.125 mm thickness were significantly different from others, with a higher error rate (mean absolute error=0.17). Measurements at 0.5mm thickness showed a significant difference with those at 0.125 and 2mm (mean absolute error=0.15). Measurements at 2mm thickness were significantly different from those at 0.125 mm thickness (mean absolute error=0.13).The average error rate was lower in 2mm thickness and the measurements were more accurate.Conclusion: A statistically significant difference was seam between CBCT measurements and actual sizes in different slice thicknesses. The differences were below 1mm, and clinically acceptable.