Predicting the Number of Fibular Segments to Reconstruct Mandibular Defects.

Abstract

Several classification schemes have been proposed to categorize mandibular defects following surgical resection; however, there is a paucity of data to guide an optimal reconstruction. This study examines the feasibility of using a geometric algorithm to simplify and determine the optimal reconstruction for a given mandibular defect. This algorithm is then applied to three different mandible defect classification schemes to correlate the defect type and number of bony segments required for reconstruction. Computed tomography (CT) scans of 48 mandibles were decomposed into curvilinear representations and analyzed using the Ramer-Douglas-Peucker algorithm. In total, 720 mandibular defects were created and subsequently analyzed utilizing three commonly referenced classification systems. For each defect, the number of bony segments required to reconstruct each defect was computed. A wide variance in the number of segments needed for optimal reconstruction was observed across existing classifications. A six-segment total mandible reconstruction best reconstituted mandibular form in all 48 mandibles. Defect classification schemes are not adaptable to predicting the number of fibula segments required for a given defect. Additionally, cephalometric templates may not be applicable in all clinical settings. The Ramer-Douglas-Peucker algorithm is well suited for providing case-specific predictions of reconstruction plans in a reproducible manner. IV Laryngoscope, 130:E619-E624, 2020.