Development of a Computer Algorithm for Pedicle Screw Placement

In the surgical treatment of spine instabilities or deformities, instrumentation with pedicle screws is used to obtain immediate and permanent stability. The planning relies on surgeon experience and in this study we have shown the development of a tool created to assist surgical planning. It elaborates CT data commonly available for most patients undergoing spinal surgery. It has in input parameters selected from the surgeon and supplies in output feasible screw trajectories, cortical bone detected, and amplitude of the entry region. We found that within the identified safe screw paths, insertion angle can be adjusted to increase cortical purchase without negative influence on screw pullout strength. Few authors have suggested to select screw in diameter equivalent to 80% of the pedicle transverse width, but a common accepted criteria for diameter selection has not been proposed yet. This value was found associated with reduced incidence of pedicle breaches and highest values of pullout loads within the cadaveric experiments performed. When we attempted to generalize the relationship between these variables, we found a weak linear correlation between pullout load and SD/PW ratio. With exemption for middle thoracic spine levels, for which available hardware is limited in sizes, the research highlighted the importance of cortical purchase. We developed a finite element model that closely replicate the failure load seen in the cadaveric experiments to further investigate the correlation between cortical bone purchase, as calculated in the algorithm, and pullout strength. We simulated four different screw diameters and found that the volume of detected cortical bone is correlated to pullout strength. In the end, we suggested trajectories with increased cortical purchase calculated using the algorithm to a surgeon trained for straightforward screw placement. The surgeon implanted the screws with same accuracy that despite the different dimensions revealed same mechanical performances. In conclusion, pedicle screw placement is a complex procedure that mainly relies on surgeon experience. We created an algorithm that assists the surgical planning elaborating CT data already in use for intraoperative navigation. The algorithm calculates parameters of proven clinical relevance and has shown the potential to easy screw selection, planning and intraoperative placement.