Comparative analysis of dental parameters within 3D cephalometric analysis using artificial intelligence

This study aimed to evaluate the comparability of three approaches to 3D cephalometric analysis: manual tracing in Invivo (Human Invivo), automated tracing using Diagnocat AI, and AI-assisted cephalometry in Invivo (AI Invivo). Materials and methods. A total of 30 CBCT scans were analyzed retrospectively, and measurements included overjet, overbite, incisor inclination, and interincisal angles. Statistical analysis comprised descriptive statistics, normality testing, ANOVA or Kruskal–Wallis tests with post-hoc comparisons, and intraclass correlation coefficient (ICC) evaluation. Pairwise differences were interpreted relative to pooled standard deviation (SD): <1 SD indicated comparability, 1–2 SD a moderate deviation, and ≥2 SD a large deviation.Results demonstrated that for the majority of parameters, differences across all three methods were below one SD, confirming high comparability and reproducibility. No parameters exceeded the 2 SD threshold. Diagnocat AI provided clinically acceptable outcomes while offering practical advantages such as reduced operator variability, faster processing time, and lower cost compared to manual tracing and AI Invivo. Conclusion. These findings suggest that both Diagnocat AI and AI Invivo may serve as reliable alternatives or adjuncts to manual cephalometry in orthodontic practice.