Artificial Intelligence in the Surgery-First Approach: Harnessing Deep Learning for Enhanced Condylar Reshaping...

Area of Science:

• Orthognathic Surgery
• Artificial Intelligence in Medicine
• Dental Imaging Analysis

Background:

• The surgery-first approach (SFA) reduces orthognathic surgery treatment time by omitting pre-surgical orthodontics.
• SFA's compromised occlusion may pose risks of condylar displacement and remodeling.
• Comparing SFA and surgery-late approach (SLA) outcomes for condylar behavior is crucial.

Purpose of the Study:

• To analyze condylar behavior using AI and deep learning.
• To compare SFA and SLA outcomes regarding condylar morphology.
• To evaluate the safety and efficacy of SFA with AI integration.

Main Methods:

• Retrospective analysis of 77 patients (18 SFA, 59 SLA) from 2016-2022.
• AI-powered 3D Slicer with CNN for automated CBCT scan segmentation (7h reduced to 5 min).
• Comparison of pre- and postoperative 3D models for condylar linear and rotational deviations.

Main Results:

• Both SFA and SLA demonstrated high surgical accuracy (<2 mm linear, <2° rotational deviation).
• SFA and SLA showed similar rates of condylar surface remodeling.
• Mean surface changes were 0.41 mm (SFA) and 0.36 mm (SLA), with minor differences across classifications.

Conclusions:

• Deep learning facilitates rapid, precise CBCT analysis for early condylar change detection.
• SFA does not adversely affect condylar morphology compared to SLA.
• SFA is safe and effective when integrated with AI technologies for orthognathic surgery.