Intraoperative 3D Imaging in Spine Surgery: Evaluation of Image Quality of a Robotic Flat Panel 3D C-Arm and Development of a Standard Operating Procedure

Abstract

Introduction Intraoperative 3D imaging in spine surgery is used for navigation and immediate intraoperative control.1 Major limitation is the image quality, and postoperative computed tomography (CT) remains the standard.2 Another concern is the exposure to radiation. Objective Aim of this study was to determine the effective dose of different imaging protocols of a robotic 3D flat panel C-arm in comparison to CT and to compare the image quality of these devices as well as to develop a standard operating procedure for this new technology. Materials and Methods Dose measurements were performed with a Rando-Alderson-Phantom equipped with thermoluminescense dosimeters (TLD). The phantom was exposed to different scanning protocols of the 3D C-arm (Artis zeego, Siemens, Erlangen, Germany) (high definition [3Dhd], high definition with collimation [3Dhdc], low dose protocol with collimation [3Dldc]) and CT (routine [CTr]) and low-dose protocol (CTl). For each technique, images were obtained of the thoracic spine (focus: Th6) and the thoracolumbar junction (focus: L1). Pedicle screws (titan vs. stainless steel) were inserted in a fresh swine cadaver (Th 5–7, Th 12–L2). Images were obtained with the same scanning protocols. Image quality (Likert scale) as well as the intra- and interobserver reliability regarding the screw accuracy were determined by two viewers blinded to acquisition technique and screw material. Results At the thoracolumbar junction, the effective dose was comparable for 3Dhd (4.4 mSv), 3Dhdc (4.3 mSv), and CTr (5 mSv) and CTl (4.0 mSv). A relevant reduction was achieved with the 3D low dose protocol (1.0 mSv). Focusing on Th6, a similar reduction with the 3Dldc protocol was achieved (3Dhd: 1.1 mSv, 3Dhdc: 0.8 mSv, CTr: 2.3 mSv, CTl 1.8 mSv vs. 3Dldc: 0.2 mSv). The image quality of the CT protocols were rated as “good” (Likert 4/5) by all viewers independent of the screw material. The intra- and interobserver reliability was excellent for all protocols (Spearman nonparametric correlation; r > 0.9, p < 0.001). The image quality of the 3D protocols using titan screws was rated as “good” by all viewers with excellent correlation ( r > 0.9, p < 0.001). At the thoracic spine, image quality of steel screws was rated “average” (borderline clinical use due to artifacts; Likert 3/5; r > 0.9, p < 0.001). Conclusion Modern intraoperative 3D C-arms produce images of CT-like quality, allowing immediate control of implants. Even with low-dose protocols a sufficient assessment of screw position is possible. Beside the image protocol, the quality depends on the screw material. Titan screws produce significant less artifacts and should be preferred with intraoperative 3D imaging. In normal weight patients low dose protocols should be applied. References Wich M, Spranger N, Ekkernkamp A. Intraoperative imaging with the ISO C(3D) [in German]. Chirurg 2004;75(10):982–987 Kotsianos D, Rock C, Euler E, et al. 3-D imaging with a mobile surgical image enhancement equipment (ISO-C-3D). Initial examples of fracture diagnosis of peripheral joints in comparison with spiral CT and conventional radiography [in German]. Unfallchirurg 2001;104(9):834–838