An Optical Tracking Approach to Computer-Assisted Surgical Navigation via Stereoscopic Vision

Abstract

Abstract Computer-assisted navigation has become a popular solution in surgical procedures where a high amount of precision is required. Current state-of-the-art methods of surgical navigation involve tracking reflective 3D marker spheres using IR stereo-scopic cameras. However, the cost of implementing such systems may not be affordable for smaller healthcare systems. In this paper, we propose that fully optical navigation has the potential to be a viable alternative to state-of-the-art reflective marker navigation. We use fiducial ArUco markers to facilitate the tracking of real-time position. Using two inexpensive cameras, we design and calibrate a stereoscopic camera to record the 3D position of an ArUco marker moving through space along a positioning platform. Additionally, we explore the possibility of using different color spaces and physical marker colors to improve the detection percentage and accuracy of markers. We identified that black-and-white ArUco markers using the Hue, Saturation, and Lightness (HSL) color space gave a positional mean error of 5.38 mm. Using the Red, Green, and Blue (RGB) color space gave the highest detection percentage for the same ArUco markers. In the future, the mean error can be reduced by increasing camera quality and by using a multi-stereoscopic camera setup.