Projected cutting guides using an augmented reality system to improve surgical margins in maxillectomies: A preclinical study

Background: Positive margins have been reported up to 80% in advanced maxillary cancers. Intraoperative navigation (IN) aims to improve margins, but provides a two-dimensional view of a registered instrument without anticipating any cutting directions, and the information is displayed in monitors outside surgical field. Augmented Reality (AR) can delineate margins while addressing the gaze-toggling drawback of IN. In a preclinical setting, we implemented preoperative-planned osteotomies needed for maxillectomies and projected this information on the surgical field using AR. We aimed to improve negative margin rates while retaining the benefits of AR. Methods: Five maxillary tumor models were built. Five fellowship-trained surgeons completed virtual unguided and AR-guided maxillectomies. Comparisons in terms of intratumoral cuts, close, adequate, and excessive distances from the tumor were performed. Differences between “ideal” cutting-plan and the AR-guided virtual osteotomies was obtained. Workload questionnaires to evaluate the technology were completed. Results: 115 virtual osteotomies were analyzed. Intra-tumoral and “close” margins were lower for the AR-assisted osteotomies (0.0% vs 1.9% p < 0.0001 and 0.8% vs 7.9% p < 0.0001). Proportion of “adequate” margins were higher in the AR simulations (25.3% vs 18.6%, p = 0.018). The AR osteotomies had high similarity with the pre-planned with interclass correlation index close to 1 in “adequate” margins 0.893 (95% CI: 0.804–0.949). Workload scores were better for AR-guided simulations for the domains of mental demand, performance, effort and frustration. Conclusion: The projector-based AR method improved margin delineation, and preoperative planning was accurately translated to the simulations. Clinical translation will aim to consolidate our preclinical findings to improve outcomes.