Analysis of simulated mandibular reconstruction using a segmental mirroring technique

Abstract

Purpose: When deforming pathology limits intraoperative plating of the mandible, three-dimensional (3D) models can be generated by digitally replacing the deformed segment of bone with an inverted segment from the contralateral unaffected mandible to adapt a reconstruction plate. The purpose of this study was to use 3D conformance analysis to evaluate the degree of accuracy of this "segmental mirroring" technique.

Methods: Using a pre-existing melanoma database (January 1, 2005-September 20, 2015), high-resolution computed tomography (CT) scans of the head and neck were obtained from patients without evidence of bony disease or defects involving the mandible. Using 3D software (Mimics, Materialise, Leuven, Belgium), each mandible was segmented based on four defect classes (Ic, II, IIc and III) of the Brown et al. (2016) classification system. An inverted, or "mirrored", image of each segment was digitally created and manually co-registered with the corresponding contralateral segment of the mandible. Conformance analysis was performed by calculating the root-mean-square (RMS) conformance distance and through evaluating 3D generated conformance maps. The primary outcome was degree of conformance. Data were analyzed using descriptive statistics and tests of statistical significance. The significance level was set at a p-value less than or equal to .05.

Results: A high degree of conformance (mean RMS < 1 mm) was observed when comparing all classes of simulated reconstruction. The closest conformance was observed for class III simulated reconstructions (mean RMS: 0.4 ± 0.2 mm). Inclusion of the condyle resulted in a reduced mean RMS conformance (class II: 0.5 ± 0.3 mm vs class IIc: 0.7 ± 0.5 mm; p = 0.01). There was no significant difference between RMS conformance distances when comparing side of simulated reconstruction. Evaluation of 3D mapping demonstrated reduced conformance with simulated reconstruction of the condyle and coronoid process.

Conclusion: The segmental mirroring technique can be used reliably to generate highly accurate three-dimensional models that may assist with mandibular reconstruction in circumstances where bony deformity limits intraoperative adaptation of a reconstruction plate. This technique is less accurate where pathology involves the mandibular condyle and, to a lesser degree, the coronoid process.

Keywords: 3D modeling; Computer simulation; Mandibular reconstruction.