84 Chapter 4 CBCT scanning CBCT scans were acquired using a NewTom 5G CBCT scanner (QR systems, Verona, Italy) at 110 kV, 4 mA, 0.3 mm voxel size and exposure time of 3.6 s. Each cadaver was covered with a plastic bag and placed inside the scanner as described in the users’ guide. The selected field of view was 12 cm × 8 cm. After selecting patient scan protocol, a regular scan (scanning time 18 s) with a boosted dose was initiated. The scans were saved as Digital Imaging and Communications in Medicine (DICOM) files. CBCT analysis The DICOM files were reconstructed using Matlab R2019a (Mathworks, Natick, MA, USA). Reconstruction involved segmentation at − 300 Hounsfield units (HU) for soft tissue and 350 HU for bone, filtering with a small smoothing kernel, a morphological closing and conversion to stereolithography (STL) format. Morphological closing is an operation on binary images to remove small gaps while preserving the overall shape and size. In our case, this operation was used to fill small air bubbles in the cadaveric tissue. Subsequently, the STL objects were analyzed in Meshmixer (Autodesk, San Rafael, CA, USA) independently by two researchers (ZA and CK). This analysis involved manual separation of the intra-oral cavity into four regions (Fig. 2): (I) the hard palate, bounded anteriorly and laterally by the maxillary alveolar ridge and posteriorly by the bony pterygoid hamuli. The bony reconstruction was used to determine the positions of the pterygoid hamulus, (II) the tongue, bordered anteriorly and laterally by the mandibular alveolar ridge. Posteriorly, the tongue was limited to the alveolar ridge on the sides and medially to the top view projection of the bony pterygoid hamulus, (III) the hard tissue region was defined as the total of all crowns in situ and dental alveoli of extracted teeth, and (IV) the remaining soft tissue was classified as mucosa, anteriorly limited by the crease of the lip retractor. After segmentation, the surface areas (in cm2) of the four separate regions were determined in Meshmixer.
RkJQdWJsaXNoZXIy MTk4NDMw