Fig. 12 (a) shows the Z_eff images of the multi-energy CT phantom obtained by the PCCT and DECT. The Z_eff image generated by DECT has been applied a noise reduction filter recommended by the manufacturer, so the noise is relatively lower. The Z_eff image obtained by PCCT is a filterless image, and it has the potential to generate an image having smaller noise by applying image processing filters. The quantitative analysis results are shown in Fig. 12 (b), and the mean and standard deviation (SD) values within the ROI are plotted. The Z_eff values obtained with both scanners were in good agreement with the calculated Z_eff values. The lower graphs show the residual error from the Y=X line. Based on the observed residual errors, the systematic uncertainties of Z_eff images acquired by PCCT and DECT were determined to be 0.40 and 0.20, respectively.

Fig 12: Results of verification experiment using a multi-energy CT phantom. Our algorithm analyzed the effective atomic number values correctly. The systematic uncertainty of each scanner was determined based on the observed residual error.

Fig. 13 shows the results of a phantom experiment using dental inlay-materials. A dental material was implanted in the 6th tooth on the lower right. The VMIs at 70 keV are presented as the conventional images. For images without artifacts, the dental materials could be clearly discriminated in Z_eff images presented in the lower row.

Fig 13: Axial images of a dental head phantom acquired by PCCT. The tooth indicated by the arrow was treated in which an artificial material was used on the surface. The effective atomic number image can provide valuable information for identifying the dental materials.

The Z_eff images of dental materials obtained by the experiment using the dental head phantom were quantitatively compared as shown in Fig. 14. The mean Z_eff values were plotted which were determined within the ROIs. The error bars of each data were determined taking into consideration the deviation within the ROI and systematic uncertainty. It was found that the results between PCCT and DECT were consistent.

Fig 14: Quantitative comparison of effective atomic number values obtained by PCCT and DECT. The calculation results of effective atomic number between PCCT and DECT were consistent.

Fig. 15 shows an example of a multi-planar reconstruction (MPR) image of a dental head phantom acquired by PCCT. Coronal and sagittal images of dental head phantom embedded with a hybrid CR were reconstructed for both (a) VMI and (b) Z_eff images. No significant artifacts were observed in any cross-section of the Z_eff images. This demonstrates that our method is applicable to the analysis of three-dimensional volume data.

Fig 15: Multi-planar reconstruction (MPR) image when hybrid CR was embedded in a dental head phantom. No significant artifacts were observed in the coronal and sagittal sections of the effective atomic number images.

Fig. 16 shows the procedure for applying our method to human identification in forensic dentistry. For diagnosing (a) unidentified human remains, we recommend to use (b) PCCT, if available. The PCCT produces (c) a conventional (qualitative) image, therefore a traditional diagnostic process can be performed. By integrating various information, (d) the identification is performed. Our proposal has the great advantage that if an unidentified person was examined by (b) PCCT, (e) Z_eff analysis can later be performed if necessary. When identifying the unidentified person in step (d), the dental material used in his/her treatment can be acquired from the treatment history information. If the same dental material is obtained from the manufacturer, (f) material analysis is performed based on Z_eff. The most important point of our proposal is that any equipment can be used to analyze the Z_eff value in step (f). Since the Z_eff value is a physical quantity specific to the object, it is possible to (g) confirm the quantitative analysis based on the information in (e) and (f) without depending on the determination procedure.

Fig 16: Identification procedure of unidentified person using quantitative information. If the unidentified remains are examined using PCCT, quantitative analysis can be performed using effective atomic number images in addition to qualitative analysis using conventional (qualitative) images. When the teeth have artificial dental materials, the effective atomic number value can provide supplemental identification information.