Akademik Veri Yönetim Sistemi
Dental Press Journal of Orthodontics, cilt.27, sa.4, ss.222112, 2022 (Scopus)
Objective: The aim of this study was to compare the measurements performed with digital manual (DM) cephalometric analysis and automatic cephalometric analysis obtained from an online artificial intelligence (AI) platform, according to different sagittal skeletal malocclusions.
Methods: Cephalometric radiographs of 105 randomly selected individuals (mean age: 17.25 ± 1.87 years) were included in this study. Dolphin Imaging software was used for DM cephalometric analysis, and the WebCeph platform was used for AI-based cephalometric analysis. In total, 10 linear and 12 angular measurements were evaluated. The paired t-test, one-way ANOVA test, and intraclass correlation coefficient tests were used to evaluate the differences between the two methods. The level of statistical significance was set at p < 0.05.
Results: Except for SNB, NPog, U1.SN, U1.NA, L1-APog, I/I, and LLE parameters, all other parameters presented significant differences between the two methods (p < 0.05). While there was no difference (p > 0.05) in both SNA and SNB measurements between the two methods in the Class I malocclusion group, there was a difference between both methods in the Class II malocclusion group. Meanwhile, only the SNA in the Class III malocclusion group was different (p < 0.05). The ANB angle differed significantly in all three malocclusion groups. For both methods, all parameters except CoA and CoGn were found to have good correlation.
Conclusion: Although significant differences were detected in some measurements between the two cephalometric analysis methods, not all differences were clinically significant. The AI-based cephalometric analysis method needs to be developed for more specific malocclusions.
Keywords: Artificial intelligence. Automatic cephalometric analysis. Digital cephalometric analysis. Skeletal malocclusion