Automated Assessment Of Condylar Changes Following Orthognathic Surgery

Introduction and Objective: This study proposes and validates a fully automatic assessment of volume changes in the mandibular condyle following orthognathic surgery.

Material and Methods: Two sets of cone-beam computed tomography scans were included (scans with segmentation of the entire mandible and pre- and postoperative scans with segmentation of the ramal segments). With both sets, convolutional neural networks were trained using a coarse-to-fine strategy to predict mandibular condyle changes. For validation, the agreement between the proposed fully automatic assessment and a validated state-of-the-art semi-automatic method was calculated by mean difference and intra-class correlation coefficients (ICC).

Results: Forty mandibular condyles from 20 patients (16 female, 4 male, mean age 27.6 years) with class II malocclusion and maxillomandibular retrognathia, who underwent bimaxillary surgery, were assessed. The fully automated method was significantly faster than the semi-automated method (3 minutes versus 30 minutes). The difference of the condylar volume change measurements produced by the two methods was not significant (mean=2.6%, standard deviation=1.8%, p=0.06). The agreement between the two methods was excellent (ICC=0.99). An systematic underestimation of pre- and postoperative condylar volumes and condylar volume changes was observed.

Conclusion: The fully automated assessment demonstrated excellent reliability for quantification of condylar volume changes following orthognathic surgery. The short processing time allows for integration into routine clinical practice.