Repeatability of gray value-based bone density measurements in cone beam computed tomography (CBCT) images under different acquisition protocols.

Abstract

Background: Accurate and repeatable assessment of bone density is a critical factor in dental implant planning, as it influences decisions related to implant size, site preparation, and primary stability, ultimately affecting treatment outcomes and long-term osseointegration. The variability in Cone-Beam Computed Tomography (CBCT) gray values across different acquisition protocols can compromise measurement repeatability, challenging the reliability of preoperative bone assessments. This study aimed to assess the repeatability of a gray value-based bone density measurement tool along three-thirds of a virtual implant in CBCT under different acquisition protocols.

Methods: Three pork ribs (5 × 3 × 2 cm) were scanned three times each using a VistaVoxS 3D CBCT scanner under four acquisition protocols, varying field-of-view (5 × 5 and 13 × 5), kVp (98, 94, 79), mA (11, 9, 6), and voxel size (0.08 and 0.2 mm), resulting in 144 scans. A virtual implant (10 × 5.2 mm) was centered on each rib image, and gray values along the thirds of the implant were measured using AIS 3DAPP 5.0 software. Repeatability was assessed with Friedman's test, Kendall's agreement coefficient, and the chi-square test.

Results: The mean CBCT gray values in the cervical region were significantly lower regardless of the protocol (p ≤ 0.5). Individual exposures differed but showed very good agreement (0.89-0.83).

Conclusions: The bone density tool showed reliable gray value measurements across CBCT protocols, with field of view affecting repeatability more than voxel size. Significant gray value differences in the cervical region highlight the need for cautious interpretation in this area. Standardizing imaging settings is recommended to improve implant planning accuracy.