Development and Validation of an Automatic Pre-Surgical Planning Tool for Long Bone Osteotomy

Abstract

Three-dimensional (3D) computer-assisted planning is increasingly used for long bone osteotomies. However, creating an optimal pre-surgical plan remains challenging due to the procedure’s many degrees of freedom, leading to high human workload and associated costs. To address this, we present a fast, interactive surgical planning tool that maximises bone contact and minimises bone protrusion, while favouring wedge-type corrections. It employs an evolutionary multi-objective optimisation algorithm to generate a set of Pareto-optimal osteotomy solutions. This allows users to choose an optimal trade-off solution based on clinical experience. The approach proposed in this work generates optimised plans in under fifteen minutes. It features a user interface, integrated in Siemens NX, making it readily integrable into existing pre-operative workflows. Quantitative validation on past cases showed that our tool produced solutions with better or equal objective values compared to manual plans in 10 of 12 cases, with an average increase of bone contact of 3 % (range: –0.7 % to 31 %). A blinded assessment by surgeons and experiments with 3D printed bones confirmed the clinical relevance and feasibility of the automatic plans. Incorporating this automatic osteotomy planning tool can therefore improve the quality of the selected pre-surgical plan, speed up the workflow and increase the volume of cases handled by the 3D lab.