CBCT Measurements of Alveolar Bone Height in Cadavers vs. Patients

Objectives: To compare the accuracy of alveolar bone height measurements made from cone beam computed tomography (CBCT) images of dental arches in cadavers and patients.
Methods: Cadaver heads and patients were imaged by CBCT (i-CAT 17-19, Imaging Sciences International, Hatfield, PA). Cadaver heads each had 3 scans made involving differing numbers of projection images on either side of the default setting (range 169 to 619 projection images; 0.2-0.3mm voxel size). Patients scheduled for periodontal surgery had scans made using the default setting (309 projection images; 0.3mm voxel size). Following a standardized protocol for selected teeth (intact, with no metallic restorations or discernable trauma to the alveolus), measurements of facial alveolar bone height were made from 0.5mm slice reconstructions centered on the long axis of each tooth. Direct measurements of facial bone height facial were subsequently made with the same teeth following dissection (cadavers), or during surgery (patients). Comparison between CBCT and direct measurements were calculated and the level of agreement assessed using Pearson’s correlation coefficients and Bland-Altman plots.
Results: With cadavers, mean differences, CBCT minus direct measurements, ranged from 0.17+/-0.12 to 0.41+/-0.32mm. Comparisons of CBCT and direct measurements showed high correlation coefficients (>0.97) and a relatively narrow Bland-Altman 95% level of agreement (<1.1mm). With patients, mean differences were 2.49+/-2.61mm (CBCT images showed lower bone height relative to direct measurements). Comparisons of CBCT and direct measurements showed a poor correlation coefficient (0.81) and a relatively wide Bland Altman 95% level of agreement (5.1mm).
Conclusions: With cadavers, CBCT measurements of alveolar bone height show high accuracy among scans made using varying projection images. With patients, using the same CBCT unit and matching one of the scan settings used with the cadavers, CBCT measurements were relatively inaccurate in that the presence of thin bone was often underestimated.