Transpedicular screw placement: image-guided versus lateral-view fluoroscopy: in vitro simulation

Abstract

Study design: In vitro evaluation of monitoring screw placement using an image-guided system compared with the routine use of an image intensifier.

Objectives: To compare a new computer-guided method of monitoring pedicle screw placement with a well-known method of monitoring using an image intensifier, using measurements of screw placement time and accuracy.

Summary of background data: Pedicle screw placement relies on the identification of anatomic landmarks for the point of entry of the pedicle. The direction and depth of screw penetration are guided by an intraoperative lateral-view image intensifier. The use of frameless stereotaxy for computer-aided pedicle screw placement may increase the accuracy and safety of the screw insertion. To the authors' knowledge, there are no published data comparing these systems on the basis of operative time and screw placement accuracy.

Method: Eight human cadaveric sections of five vertebrae each were used for an in vitro simulation of pedicle screw placement. Four spine surgeons were chosen to simulate the transpedicular screw placement. Each surgeon placed one screw into each pedicle of two spine sections (10 vertebrae, 20 screws). The surgeon was assisted by the lateral-view image intensifier on one spine section and by the navigational system on the second one. The four surgeons placed 80 pedicle screws. Forty screw placements were monitored by fluoroscopy and 40 by the image-guided navigational system. The time spent to place one screw was recorded, as well as the remarks by each surgeons on each method. Spines were rescanned, and the positions of the screws were compared between the group on which the image intensifier has been used and the group on which the navigational system had been used.

Results: In the image-guided technique group, one thoracic screw disrupted the lateral cortex of the pedicle, the average distance to the anterior wall of the body was 5 mm, and the average time for the insertion of one screw was 13.5 minutes. In the other group, two screws disrupted the inner cortex of a thoracic pedicle, the average distance to the anterior wall was 10.7 mm, and the average time for the insertion of one screw was 4 minutes.

Conclusions: In vitro computer-aided pedicle screw insertion is more accurate than lateral-view fluoroscopy in the thoracic spine. The main disadvantage is the time consumption compared with that required by lateral-view fluoroscopy. The total time of the surgical operation should be decreased with the future development of these techniques.