A posterior landmark-based registration-free method to identify pedicle screw trajectories for robot-based navigation: A proof-of-concept

Pedicle screw placement (PSP) is a frequent intervention in spine surgery. The use of robotic systems for PSP gains increasing attention, as they allow more accurate screw implantation. Current approaches pre-define screw trajectories in preoperative computer-tomography (CT) data and use registration for alignment with the intraoperative anatomy. In current approaches, screw trajectories are predefined in preoperative computed tomography (CT) data and adapted to intraoperative anatomy via registration. However, this approach is increasingly being questioned because the patient is exposed to radiation, the registration remains error-prone, and the screw trajectories cannot be adjusted intraoperatively. The EU Horizon 2020 project FAROS aims to develop a radiation- and registration-free robotic system for spine surgery. FAROS foresees using robotic ultrasound (US) to reconstruct a 3D model of the spine.  In this work, we present a new approach capable of performing PSP EP and trajectory planning only based on posterior surface models.