Mandibular tumors and radical oral cancer surgery often cause bone dysmorphia and defects. Most patients present with noticeable mandibular deformations, and doctors often have difficulty determining their exact mandibular morphology. In this study, a deep convolutional generative adversarial network (DCGAN) called CTGAN is proposed to complete 3D mandibular cone beam computed tomography data from CT data. After extensive training, CTGAN was tested on 6 mandibular tumor cases, resulting in 3D virtual mandibular completion. We found that CTGAN can generate mandibles with different levels and rich morphology, including positional and angular changes and local patterns. The completion results are shown as tomographic images combining generated and natural areas. The 3D generated mandibles have the anatomical morphology of the real mandibles and transition smoothly to the portions without disease, showing that CTGAN constructs mandibles with the expected patient characteristics and is suitable for mandibular morphological completion. The presented modeling principles can be applied to other areas for 3D morphological completion from medical images. Clinical trial registration: This study is not a clinical trial. Patient data were only used for testing in a virtual environment. The use of the digital data used in this study was ethically approved.

No abstract

Mandibular tumors and radical oral cancer surgery often cause bone dysmorphia and defects. Most patients present with noticeable mandibular deformations, and doctors often have difficulty determining their exact mandibular morphology. In this study, a deep convolutional generative adversarial network (DCGAN) called CTGAN is proposed to complete 3D mandibular cone beam computed tomography (CBCT) data from CT data. After extensive training CTGAN was tested on 6 mandibular tumor cases, resulting in 3D virtual mandibular completion. We found that CTGAN can generate mandibles with different levels and rich morphology, including positional and angular changes and local patterns. The completion results are shown as tomographic images combining generated and natural areas. The 3D generated mandibles have the anatomical morphology of the real mandibles and transition smoothly to the portions without disease, showing that CTGAN constructs mandibles with the expected patient characteristics and is suitable for mandibular morphological completion. The presented modeling principles can be applied to other areas for 3D morphological completion from medical images.

Aim'This study aims to establish an open-source algorithm using Python to analyze the accuracy of guided implantation, which simplifies interstudy comparisons.MethodsGiven ≥3 landmark pairs, this Tri-Point (TriP) method can register images. With ≥4 landmark pairs, TriP can calculate system errors for image registration. We selected 8 indicators from the literature. Considering development errors in new bone on cone beam computed tomography (CBCT), we added the indicators of apical rectified deviation (ARD) and coronal rectified deviation (CRD), providing accurate references but neglecting depth deviations. Our program can calculate and output these indicators. To evaluate the TriP method’s feasibility, an implantation group assisted by a Visual Direction-INdicating Guide (VDING) was analyzed. Accuracy was measured with the traditional and proposed TriP methods. Factors affecting the system error of the method were then analyzed.ResultsComparisons with paired-sample t-tests showed that our TriP method was similar to the traditional method in evaluating implantation accuracy, with no significant difference (P>0.05). The average system error was 0.30±0.10 mm when the TriP method evaluated the VDING template. The results showed that increasing the provided landmarks from 4 to 5 pairs decreased the between-group differences significantly (P<0.05). With ≥6 pairs of landmarks, the system error tended to be stable, and the groups showed no statistically significant differences (P>0.05). Large distances between landmarks are helpful in reducing system error, as demonstrated with a geometric method.ConclusionsThis study established an open-source algorithm to analyze the accuracy of guided implantation with system errors reported.