A Proof of Concept: Optimized Jawbone-Reduction Model for Mandibular Fracture Surgery

Abstract

Previous research on computer-assisted jawbone reduction for mandibular fracture surgery has only focused on the relationship between fractured sections disregarding proper dental occlusion with the maxilla. To overcome malocclusion caused by overlooking dental articulation, this study aims to provide a model for jawbone reduction based on dental occlusion. After dental landmarks and fracture sectional features are extracted, the maxilla and two mandible segments are aligned first using the extracted dental landmarks. A swarm-based optimization is subsequently performed by simultaneously observing the fracture section fitting and the dental occlusion condition. The proposed method was evaluated using jawbone data of 12 subjects with simulated and real mandibular fractures. Results showed that the optimized model achieved both accurate jawbone reduction and desired dental occlusion, which may not be possible by existing methods.

Keywords: Computer-assisted surgical planning; Dental occlusion; Jawbone reduction; Mandibular fracture.

Fig. 1

Dental landmark extraction in the maxilla. a Alignment of the maxilla with the world coordinate system. b Arch line fitting of the maxilla. c Slicing perpendicular planes. d Tooth separation. e Central groove extraction. f Maxillary dental landmarks

Fig. 2

General position of mandibular sections. a Alignment of mandible segments with the arch line of the maxilla. b Buccal cusp extraction. c Mandibular dental landmarks. d Fracture sectional features. e General positioning between maxilla and mandible segments

Fig. 3

Surface distance map between the pre-fracture jawbone and jawbone-reduction model. (p1´) Error by fracture sectional features only. (p1´´) Error by dental landmarks only. (p1–p8) Error by proposed method. The red dotted lines indicate the fracture locations

Fig. 4

Surface distance map with various fracture and teeth arrangements. a 10–4 tooth fracture, b 11–3 tooth fracture, and c two-teeth-missing situation

Fig. 5

Optimized jawbone-reduction model with real patient data

Fig. 6

Parameter tuning for the CLPSO. Each graph represents the average and variance of the objective function output with various parameter setting. The red boxes indicate the selected values. a Minimum value of the inertia weight. b Maximum value of the inertia weight. c Acceleration constant

Fig. 7

Error of the reduction model with weight setting for sample cases (participants 1–3). The green box indicates the selected weights