Reliability of Linear Measurements Made on Reconstructed CBCT Images

Abstract

Objectives: Repeatability and accuracy of measurements made on cone-beam computed tomography (CBCT) images are critical in dental practice especially in implantology. The aim of this study was to investigate the reproducibility of linear measurements made on reconstructed CBCT images. Methods In this in vitro, experimental study, 5 radiopaque markers were attached to the molars (left and right side), premolars (left and right side) and midline areas of 10 human cadaver dry mandibles. The distance between the markers and the lower border of the mandible was measured by a digital caliper and considered as the gold standard. CBCT images were taken, and the distance between the markers and the lower mandibular border was measured on cross-sectional images by three maxillofacial radiologists using Romexis software. The same measurements were made 1 month later to assess the reproducibility of measurements. The intra-class correlation coefficient was calculated to assess the repeatability and agreement between the observers. Results: Compared with the gold standard, the mean error percentage in linear measurements was calculated to be 3.25%. The overall reproducibility of CBCT linear measurements was 0.865. The inter-observer agreement was calculated to be 0.972. Conclusion: CBCT showed acceptable accuracy, repeatability and reliability for linear measurements, and can be used as an accurate tool for this purpose.