Material and methods

CBCT data sets acquired from 30 patients were used in this study. Some subjects with pathologic conditions were included, including 4 who were being evaluated for temporomandibular disorder, 1 with a pneumocoele of the maxillary right antrum, and 1 with an odontogenic cyst surrounding an impacted mandibular third molar. Subjects with developmental conditions such as impacted teeth and orofacial clefts were also permitted ( Table I ). Selection to homogenize the sample with respect to age, sex, and ethnicity was not specifically performed. The sample included 13 men and 17 women with an average age of 22 years 8months±12 years 1month (median, 19 years 2month). The age was not recorded in the database for 1 patient, who was estimated to be 25 years 0months at the scan. Most subjects were white (n = 17), followed by Asians (n = 6), Hispanics (n = 5), and blacks (n = 2). The most common reasons for referral for the CBCT scan were impacted teeth (10 subjects) and evaluations of the temporomandibular joints for disorders (8 subjects).

CBCT data from 2 systems were used. Initially, imaging data were acquired from the NewTom QR DVT 9000 (Aperio, Sarasota, Fla). This was later followed by the CB MercuRay (Hitachi Medico Technology, Tokyo, Japan). For inclusion in this study, scans had to include completely the portion of the maxilla at and below the nasal floor (anterior nasal spine-posterior nasal spine level), the entire mandible, and the upper airway at least as superiorly as the inferior turbinate and as inferiorly as the epiglottis. CBCT data from each subject were segmented by using CBWorks software (version1.0, CyberMed, Seoul, Korea) to provide shaded surface meshes representing the 3D structure of each subject’s mandible. In the process, information regarding the subject’s identity was removed.

For our studies, segmentation was achieved first by selecting a volume of interest as close to the structure of interest as possible. Then, the object was threshold-segmented and slice-edited by hand to remove any visible extraneous scatter, artifacts, or background, similar to the method of Meehan et al. For areas with widely different threshold ranges because of regional differences in bone density—eg, the body and ramus of the mandible vs the mandibular condylar head—each region was segmented separately. Then threshold settings were removed, and the 2 data sets were added together by using a Boolean operations module. After segmentation, the resulting set of voxels, or “point cloud,” was transformed by the software into a 3D surface composed of a tetrahedral mesh with vertices or nodes related to voxel positions. For this project, surface quality was set at high, the critical surface angle was left at the default of 120°, and unconstrained smoothing was used.

CBCT data from each subject were segmented with CBWorks software to provide shaded surface meshes representing the 3D structure of each subject’s mandible. A combination of threshold segmentation and manual slice editing was used to obtain refined surfaces with few artifacts. The segmentation process involved using the sculpt tool of CBWorks to narrow the scan into a volume of interest including only the structure that was to be segmented. Threshold segmentation tools were used to isolate the osseous structures. For the mandible, the condyles (neck and head) were segmented separately from the rest of the jaw, because of differences in bone density and attenuation coefficient values of these regions. Boolean operations were then used to combine several data sets of regions of vastly different x-ray attenuation coefficients. Manual slice editing was performed on these data sets to remove most scatter and other noise artifacts, as well as to fill in unanatomic pseudo-foramen artifacts that can appear in thin or cancellous bone, such as in the condylar head and posterior palate. After segmentation, the resulting set of masks (highlighted areas representing each structure of interest in each slice of the CBCT scan) was rendered into a shaded surface mesh in the CBWorks surface shaded display tool ( Fig 1 ). The 30 resulting surface meshes were exported from CBWorks in a virtual reality modeling language (.vrl) format into Amira software (version 3.1, Mercury Computer Systems, Berlin, Germany) for further analysis.

Process of reference surface selection, beginning with a surface-rendered image to which landmarks are embedded to determine mean landmark coordinates and compare the mandibles with RMS analysis.

A set of 39 landmarks was embedded in each mandible with the Amira landmark tool ( Table II ). Landmarks were embedded pairwise in the arbitrary reference and in each structure. The Amira landmark surface warp tool was set in a rigid transformation method to align the surfaces based on the embedded landmarks. Individual surfaces were then enlarged or compressed uniformly in x-, y-, and z- dimensions by hand with the Amira transform editor dialog to scale each surface to the same size as the arbitrary reference surface. Once the surfaces were aligned and scaled, a subset of landmarks was used to identify, mathematically, the “most average” surface of the 30 mandibles, and to rank the remaining surfaces relative to it. Because the objects now had similar alignment and size, the remaining differences in the surfaces were related to differences in shape. For the mandible, the x-, y-, and z-coordinates of 7 landmarks (menton, and left and right gonion, condylion, and tips of the coronoid processes) were determined by using the query mode of the Amira landmarks tool, and the rectangular coordinate values were also established.

Stay updated, free dental videos. Join our Telegram channel

Join