OBJECTIVES: We applied the fusion function of dental analysis software to examine whether the anatomical relationships of tooth roots when compared between reconstructed post-mortem CT (PMCT) and dental radiographs can aid dental identification. METHODS: One PMCT image taken from a cadaver (43-year-old male; Cadaver 1) was compared with 64 digital dental radiographs of the left and right upper and lower molars from Cadaver 1 and 30 other cadavers. Five corresponding anatomical reference points were marked on each image. After adjusting the angle and magnification using the fusion function, the automatically calculated error in pixels was determined five times for each of four sites on the images. RESULTS: Comparison of the PMCT image with the dental radiographs from the other cadavers revealed obvious discrepancies in the anatomical positioning of the teeth. When t-tests were applied to the data from any of the four sites, the error in pixels was found to be significantly smaller (p < 0.001) between Cadaver 1 and the other cadaver images. The average error in pixels between the PMCT and dental radiographs was smaller in the lower jaw than in the upper jaw. CONCLUSIONS: This method uses corresponding reference points on two images to calculate the error between the regions that contain all points. This feature also makes it possible to compare images taken with different modalities. The demand for a dental identification method involving PMCT is likely to increase, and we expect that the accuracy of dental identification will improve by using radiological images.
OBJECTIVES: We applied the fusion function of dental analysis software to examine whether the anatomical relationships of tooth roots when compared between reconstructed post-mortem CT (PMCT) and dental radiographs can aid dental identification. METHODS: One PMCT image taken from a cadaver (43-year-old male; Cadaver 1) was compared with 64 digital dental radiographs of the left and right upper and lower molars from Cadaver 1 and 30 other cadavers. Five corresponding anatomical reference points were marked on each image. After adjusting the angle and magnification using the fusion function, the automatically calculated error in pixels was determined five times for each of four sites on the images. RESULTS: Comparison of the PMCT image with the dental radiographs from the other cadavers revealed obvious discrepancies in the anatomical positioning of the teeth. When t-tests were applied to the data from any of the four sites, the error in pixels was found to be significantly smaller (p < 0.001) between Cadaver 1 and the other cadaver images. The average error in pixels between the PMCT and dental radiographs was smaller in the lower jaw than in the upper jaw. CONCLUSIONS: This method uses corresponding reference points on two images to calculate the error between the regions that contain all points. This feature also makes it possible to compare images taken with different modalities. The demand for a dental identification method involving PMCT is likely to increase, and we expect that the accuracy of dental identification will improve by using radiological images.