Analysis of Artificial Intelligence-assisted Three-Dimensional Landmark Tracing in Orthognathic Surgery Planning: A Comparative Study

Read the full article See related articles

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

Objectives : Orthognathic surgery aims to correct dentofacial deformities by repositioning the maxillomandibular complex. The advent of digital technology, particularly Virtual Surgical Planning (VSP), has enhanced the precision and efficiency of these procedures. Despite this, comparative analyses of VSP software accuracy remain limited. This study evaluates the accuracy of VSP across multiple software platforms and examines the time-saving potential of Artificial Intelligence (AI)-assisted cephalometry. Mateials and Methods: Fifteen patients were evaluated using three methods: manual tracing with Invivo6, manual tracing with ON3D, and AI-assisted tracing with ON3D. Positional differences were measured at key landmarks (A, ANS, PNS, B, Pog, Me, RU6C, RL6C). Statistical analyses included the Wilcoxon signed-rank test and ANOVA to assess accuracy, while interclass correlation evaluated time measurement reliability among four researchers. Additionally, ANOVA compared time efficiency across methods. Results: Significant positional differences between VSP and Actual Surgical Outcome (ASO) were observed at B, Pog, and Me along the z-axis, and discrepancies were noted at RU6C and RL6C on the y and z axes between Invivo6 and ON3D. AI-assisted cephalometry (Method 3) showed a substantial reduction in analysis time. Conclusions: Mandibular discrepancies between VSP and ASO appear related to postoperative CBCT adjustments, particularly incomplete soft tissue and occlusal adaptation, whereas maxillary discrepancies were minimal (within 2 mm), supporting the accuracy of all methods in the maxilla. Differences at dental landmarks between Invivo6 and ON3D stem from the inclusion of dentition scan data. Clinical Relevance: The use of AI-assisted cephalometry greatly enhanced time efficiency while maintaining accuracy, indicating the potential for AI integration in optimizing orthognathic surgical planning workflows.

Article activity feed