Visualization-enhanced surgical navigation: Silhouette registration, optimization, and monitoring

Accurate surgical instrument monitoring plays a critical role in computer-assisted surgery. However, existing navigation systems still fall short of providing comprehensive monitoring required by surgeons. An intuitive silhouette or spatial projection of the instrument is essential for assisting surgeons in monitoring its movement direction and contact with lesion area. This paper introduces a visualization-enhanced surgical navigation for enhancing localization and visualization, providing highly robust localization and high quality silhouette monitoring. The key to overcoming these challenges lies in registering overall the surgical instrument. Our navigation system achieves this through a feature-dense position-sensing marker, silhouette-based reconstruction method, and density field collapsing optimization method. In the experiment, the comprehensive geometric reconstruction accuracy of our method for surgical instrument achieves 0.161 ± 0.152 mm. The comprehensive localization accuracy reaches 0.739 ± 0.182 mm and 1.429 ± 1.325°. These results demonstrate that the proposed navigation system provides precise and reliable surgical instrument monitoring, enabling surgeons to better observe and monitor the instruments within the lesion area during surgery.