Evaluation of a novel algorithm for automating virtual surgical planning in mandibular reconstruction using fibula flaps.

Virtual surgical planning plays an increasingly important role in jaw reconstruction. The aim of the present study was the evaluation of the clinical applicability of a novel algorithm for automating virtual mandibular reconstruction using fibula flaps. The software uses Computed-Tomography of the facial skeleton and the lower leg of 63 subjects, implemented in Python programming language. The developed algorithm is based on individual bone curvatures of the mandible and fibula. Ten different defects were generated for each mandible by virtually defined cutting planes. Three experienced surgeons reviewed all reconstruction proposals generated by the algorithm according to a visual analogue scale. The possible correlation between the ratings and the prioritization of the algorithm and the calculation time for the reconstructions were analyzed. Spearman analysis showed a strong correlation -0.613 (p < 0.001) between the deviation of the reconstruction result from the target line and the average assessment of the surgeons as well as a moderate correlation -0.448 (p = 0.013) between surgeons' assessments and the prioritization by the algorithm. The calculation time for twenty reconstructions per defect took between 4.99 s and 483.5 s depending on defect size and location. The evaluated algorithm automatically creates valid reconstruction results with acceptable computation time, which have received a high level of confirmation from experienced surgeons.

Introduction of an algorithm for planning of autologous fibular transfer in mandibular reconstruction based on individual bone curvatures.

BACKGROUND: Reconstruction of the mandible with autologous transplants is a challenging task and current computer-aided surgical planning remains cumbersome. Thus, the aim of the present study was to create an automated computational approach for this procedure. METHODS: The developed algorithm is based on curves following characteristic anatomical features. Geometrical data from a physiological mandible and a fibula were used to generate six different defects, and geometrical accordance was investigated to demonstrate the applicability of the method with different reconstruction parameters (n = 309). RESULTS: The method proved to be applicable, it recognized given clinical constraints and the values of accordance could be used to quantify the success of reconstructions. CONCLUSIONS: With the present approach, the complex three-dimensional task of mandibular reconstruction was simplified, and thus it allows implementation in clinical routine. The computational planning that is proposed may be used to design cutting guides or as geometrical input data for real-time navigated surgery.