Sensor-fusion based augmented-reality surgical navigation system

Abstract

Surgical navigation systems can overlay medical images on the patients and help surgeons obtain surgical information in operations through motion tracking and fusion display. To improve the performance of display and accuracy of motion tracking of surgical navigation systems, this paper proposes an inertial-optical hybrid tracking system and a half-mirror based surgical navigation system. Extended Kalman filter is used to fuse the measurements from the inertial and optical sensors, as well as provide the orientation and position of the surgical tool and patient. The hybrid tracking method can compensate for the occlusion under the conditions that the markers are partially occluded in long-term. The surgical navigation image is related to the real targets by the hybrid tracking method and displayed in a half-mirror display device which overlays the virtual targets on the surgical tool and patient. The experimental results indicate that the proposed surgical navigation system based on sensor fusion can accurately track the motion of the surgical targets and display the guiding image in real time.