Volumetric characterization of the Aurora magnetic tracker system for image-guided transorbital endoscopic procedures.

In some medical procedures, it is difficult or impossible to maintain a line of sight for a guidance system. For such applications, people have begun to use electromagnetic trackers. Before a localizer can be effectively used for an image-guided procedure, a characterization of the localizer is required. The purpose of this work is to perform a volumetric characterization of the fiducial localization error (FLE) in the working volume of the Aurora magnetic tracker by sampling the magnetic field using a tomographic grid. Since the Aurora magnetic tracker will be used for image-guided transorbital procedures we chose a working volume that was close to the average size of the human head. A Plexiglass grid phantom was constructed and used for the characterization of the Aurora magnetic tracker. A volumetric map of the magnetic space was performed by moving the flat Plexiglass phantom up in increments of 38.4 mm from 9.6 mm to 201.6 mm. The relative spatial and the random FLE were then calculated. Since the target of our endoscopic guidance is the orbital space behind the optic nerve, the maximum distance between the field generator and the sensor was calculated depending on the placement of the field generator from the skull. For the different field generator placements we found the average random FLE to be less than 0.06 mm for the 6D probe and 0.2 mm for the 5D probe. We also observed an average relative spatial FLE of less than 0.7 mm for the 6D probe and 1.3 mm for the 5D probe. We observed that the error increased as the distance between the field generator and the sensor increased. We also observed a minimum error occurring between 48 mm and 86 mm from the base of the tracker.