This study was approved by an institutional review board of our university (D2016-061).

Among the patients whose mandibles were examined by CBCT at our dental hospital between April 2012 and August 2016, 96 patients who fulfilled the following two conditions were selected.

On CBCT imaging:

The smallest field of view (FOV) of the device, 40 × 40 mm, was used.

The mental foramen and the mandibular body over a range of 30 mm or more just distal to the foramen were imaged.

Of those, 10 patients were excluded because the mandibular canals were affected by lesions on the images. The remaining 86 patients (31 male and 55 female; mean age, 55 years; age range, 19–79 years) were included in this study. The reasons for the CBCT study were to assess a dental lesion in 56 (periapical lesion in 51, root fracture in 4, and periodontal disease in 1) and treatment planning for dental implants in 30 patients.

CBCT images were obtained using 3D Accuitomo FPD (Morita Corp., Kyoto, Japan) operated at tube voltage of 87–90 kV, tube current of 5–8 mA, and scan time of 9 or 18 s. In all cases, the smallest FOV, 40 × 40 mm, was used and the images were reconstructed with a voxel size of 0.08 mm.

Using OsiriX software version 3.8.1 (http://www.osirix-viewer.com), cross-sectional CBCT images of the mandible with 1-mm thickness and at 1-mm intervals were reformatted. After the mental foramen was localized, cross-sectional images in a range of 30 mm just distal to the foramen were used for the evaluation. The range was divided into three areas, each of which was 10 mm in length. These were designated as area 1, area 2, and area 3, from anterior to posterior. Each area contained 10 cross-sectional images (Fig. 1).

Two observers (A.T. and H.I., with over 20 years’ and 3 years’ experience as oral radiologists, respectively) independently evaluated the images in a darkened room for the presence or absence of visualization of the superior and inferior walls of the mandibular canal in each of the 10 cross-sectional images in all three areas (Fig. 2). For the purpose of calibration, training was held using typical images prior to the evaluation. Each observer was blind to the other’s results. When disagreement existed between the two observers, another observer (T.K., with over 30 years’ experience as oral radiologist) made a final judgment. After the evaluation, the visibility ratio of the superior and inferior walls in each area was determined as follows: