[Effectiveness assessment of 3-D cone beam CT used in human bite marks identification]

The present study was aimed to use the 3-D cone beam CT (CBCT) as a new method in human bite marks identification which was carried out in experimental pigskin to assess its effectiveness in our laboratory. Bite marks were digital photographed according to American Board of Forensic Odontology (ABFO) guidelines. In this study, the data of the suspect's dental casts were collected by scanning in two ways: one was after plate scanning, in which the comparison overlays were generated by Adobe Photoshop 8.0 software; the other was by CBCT, which generated comparison overlays automatically. The bite marks were blind identified with the two kinds of data of the suspect's dental casts respectively. ROC curve was used to analyze the sensitivity, specificity, and 95% confidence interval. The results showed that CBCT method got a larger area under the ROC curve: 0.784 (SE = 0.074, 95% CI = 0.639-0.929), and got a very high specificity (specificity 98.7%, 95% CI = 94.5%-99.8%). Thus, this study illustrates that the CBCT used in bite mark identification is an effective and accurate tool and has stronger ability to exclude suspects compared with the conventional method, but the comparison process needs further study to enhance its effectiveness in bite mark identification.