Comparison between cephalometric measurements using digital manual and web-based artificial intelligence...

Area of Science:

• Orthodontics and Dentofacial Orthopedics
• Medical Imaging Analysis
• Artificial Intelligence in Healthcare

Background:

• Cephalometric analysis is crucial for diagnosing skeletal malocclusions.
• Digital manual (DM) analysis is a traditional method, while AI-based analysis offers potential for automation.
• Comparing these methods is essential to understand their clinical applicability.

Purpose of the Study:

• To compare cephalometric measurements between DM analysis and an AI platform across different sagittal skeletal malocclusions.
• To evaluate the reliability and differences of AI-based cephalometric analysis versus traditional DM methods.

Main Methods:

• 105 individuals' cephalometric radiographs were analyzed using Dolphin Imaging (DM) and WebCeph (AI).
• 10 linear and 12 angular measurements were assessed.
• Statistical analysis included paired t-test, ANOVA, and intraclass correlation coefficient (ICC).

Main Results:

• Significant differences were found in most parameters between DM and AI methods (p<0.05), except for specific measurements like SNB and U1.NA.
• Differences varied across malocclusion classes; ANB angle differed significantly in all groups.
• Most parameters showed good correlation between the two methods, excluding CoA and CoGn.

Conclusions:

• While AI cephalometric analysis shows promise, significant measurement differences exist compared to DM methods.
• Not all detected differences were clinically significant, suggesting potential for AI use.
• Further development of AI cephalometric analysis is recommended for improved accuracy in specific malocclusions.