Objective: The purpose of this study is to evaluate the reliability of a macroscopic personal identification test based on reconstructed panoramic radiographs obtained with dental cone beam computed tomography （dental-CBCT） and standard panoramic radiographs.
Materials and methods: We selected 100 consecutive patients who underwent dental-CBCT and panoramic radiography. However, we excluded 30 patients who had tumors and/or cysts in the maxillofacial region or in whom not all teeth were located in the field of view. Thus, 70 patients were chosen and randomly allocated to three groups （20 patients per group）. Reconstructed panoramic radiographs were produced using the ray-summation technique without reference to any panoramic radiographs or personal information. Both the reconstructed panoramic radiographs and panoramic radiographs were output to 300 dpi films. The evaluators were four dentists with 35 years, 30 years, 21 years, and 3 years’ experience of diagnostic imaging, respectively. Using the Kappa coefficient, we examined inter-rater reliability during macroscopic comparisons between the 20 reconstructed panoramic radiographs and 20 panoramic radiographs in each group （in the absence of any other personal information）. Personal identification tests of each group were conducted after an interval of at least one month.
Results: All evaluators showed a very high concordance rate （100％） within the three groups. Furthermore, when the inter-rater concordance rate for the four evaluators who examined the three groups was investigated, the Kappa coefficient was found to be 1, indicating ‘almost perfect’ reliability. In addition, changes in dental condition （the numbers of residual teeth, treated teeth, and dental implant） did not have any effect on the evaluators’ results.
Conclusion: This study showed that a macroscopic personal identification test based on the use of reconstructed panoramic radiographs derived from dental-CBCT and standard panoramic radiographs exhibited high reliability and validity.
抄録全体を表示