Image distortion and spatial resolution of a commercially available cone-beam computed tomography machine

Abstract

Introduction: Our objective was to evaluate images produced by a commercially available cone-beam computed tomography (CBCT) machine (i-CAT model 9140-0035-000C, Imaging Sciences International, Hatfield, Pa) for measurement and spatial resolution (ie, the ability to separate 2 objects in close proximity in the image) for all settings and in all dimensions.

Methods: A custom phantom containing 0.3 mm diameter chromium metal markers approximately 5 mm apart in 3 planes of space was developed for analyzing distortion and measurement accuracy. This phantom was scanned in the CBCT machine by using all 12 commercially available settings. The distance between the markers was measured 3 times on the 3-dimensional images by using a Digital Imaging and Communications in Medicine (DICOM) viewer and was also measured 3 times directly on the phantom with a fine-tipped digital caliper. A line-pair phantom was used to evaluate spatial resolution. Thirty evaluators analyzed images and assigned a resolution from 0.2 to 1.6 mm according to the separation of the line pairs.

Results: There were no statistically significant differences among the 3-dimensional images for any setting, in any dimension, or in images divided by thirds in terms of measurement accuracy. Comparison of the CBCT measurements to the direct digital caliper measurements showed a statistically significant difference (P <0.01). However, the absolute difference was <0.1 mm and is probably not clinically significant for most applications. The worst spatial resolution found was 0.86 mm. Spatial resolution was lower at faster scan times and larger voxel sizes.

Conclusions: This CBCT machine has clinically accurate measurements and acceptable resolution.