Test–retest reliability of morphological measurements of the mandible on cone-beam computed tomography-synthesized cephalograms

Abstract

Background/purposeThe current study aimed to determine the intra- and inter-rater, and intersession reliability of the determination of the morphological parameters of the human mandible using cone-beam computed tomography (CBCT)-synthesized cephalometric radiographs. Materials and methodsCBCT data of 12 mandibles were obtained and used to generate synthetic cephalograms via a digitally reconstructed radiography (DRR) technique. Eleven landmarks describing the key morphological features of the mandible on each DRR-synthesized cephalogram were identified manually six times by one senior and one junior dentist. The operation was repeated 5 days later. Twelve parameters based on interlandmark line segments and their angles were calculated. Test–retest reliability was assessed in terms of intraclass correlation coefficient (ICC) and coefficient of variation (CV) using a two-way mixed-effects model. The paired-sample t-test was used to compare differences between examiners and sessions. A one-sample t-test was employed to assess whether the difference between the examiners was significantly different from zero. ResultsVery good intrarater (senior: ICC > 0.93; junior: ICC = 0.78 for CdP-GoP, ICC > 0.91 for other parameters), inter-rater (ICC = 0.62 for CdP-GoP, ICC > 0.84 for other parameters) and intersession reliability (ICC > 0.84 for all parameters and examiners; ICC = 0.74 for CdP-GoP for junior examiner) in measuring mandibular morphological parameters were found. ConclusionThese results suggest that very good reliability could be achieved via manual identification of the anatomical landmarks without the effects of factors such as malpositioning of the head during imaging. Further investigations using the current DRR-based approach will be needed to evaluate the individual effects of these other factors on the morphological measurements.