Accuracy and reliability of different cone beam computed tomography (CBCT) devices for structural analysis of alveolar bone in comparison with multislice CT and micro-CT

Abstract

Objective: The aim of this study was to assess whether cone beam computed tomography (CBCT) may be used for clinically reliable alveolar bone quality assessment in comparison to its clinical alternatives, multislice computed tomography and the gold standard (micro-CT). Materials and methods: Six dentate mandibular bone samples were scanned with seven CBCT devices (ProMax 3D Max, NewTom GiANO, Cranex 3D, 3D Accuitomo 170, Carestream 9300, Scanora 3D, I-CAT Next generation), one micro-CT scanner (SkyScan 1174) and one MSCT machine (Somatom Definition Flash) using two protocols (standard and high-resolution). MSCT and CBCT images were automatically spatially aligned on the micro-CT scan of the corresponding sample. A volume of interest was manually delineated on the micro-CT image and overlaid on the other scanning devices. Alveolar bone structures were automatically extracted using the adaptive thresholding algorithm. Based on the resulting binary images, an automatic 3D morphometric quantification was performed in a CT-Analyser (Bruker, Kontich, Belgium). The reliability and measurement errors were calculated for each modality compared to the gold standard micro-CT Results: Both MSCT and CBCT were associated with a clinically and statistically (P < 0.05) significant measurement error. Bone quantity-related morphometric indices (bone volume fraction 8.41 % min to 17.90% max, bone surface density-0.47 mm(-1 min) to 0.16 mm(-1 max) and trabecular thickness 0.15 mm min to 0.31 mm max) were significantly (P < 0.05) overestimated, resulting in significantly (P < 0.05) closer trabecular pores (total porosity percentage -8.41% (min) to -17.90% (max) and fractal dimension 0.08 (min) to 0.17 (max)) in all scanners compared to micro-CT. However, the structural pattern of the alveolar bone remained similar compared to that of the micro-CT for the Pro(Max) 3D (Max), NewTom GiANO, Cranex 3D, 3D Accuitomo 170 and Carestream 9300. On the other hand, the Scanora 3D, i-CAT Next Generation, standard and high-resolution MSCT displayed an overrated bone quantity and aberrant structural pattern compared to other scanning devices. The calculation of morphometric indices had an overall high reliability (intraclass correlation coefficient [ICC] 0.62 (min) to 0.99 (max)), except for the i-CAT Next Generation CBCT (ICC 0.26 (min) to 0.86 (max)) and standard resolution MSCT (ICC 0.10 (min) to 0.62 (max)). Conclusions: This study demonstrated that most CBCT machines may be able to quantitatively assess alveolar bone quality, with a level of accuracy and reliability that approaches micro-CT. One may therefore propose to extrapolate this to clinical CBCT imaging, certainly when there is a need for implant rehabilitation in dentate jaw bones.