Evaluation of pedicle screw position on computerized tomography scans

Abstract

The quality of a computerized tomography (CT) scan is significantly reduced by metal artifact caused by a pedicle screw system. The purpose of this study was to develop a method of facilitating the evaluation of pedicle screw position on CT scans obtained after screw insertion. The authors developed an algorithm to process spiral CT scans in a personal computer. This uses a digital image enhancement technique, the curve change–based intensity transformation algorithm. This method can generate a clear image of the screw outlines while reducing metal artifact. The resulting images are displayed in arbitrary planes as well as in axial, coronal, and sagittal planes, to support better the evaluation of pedicle screw position. The algorithm was tested using CT scans obtained in 37 patients in whom 186 pedicle screws had been placed. There were five types of screw systems, all of which were made of titanium alloys. In all cases algorithm-based determination of screw position became more convenient and more accurate than when using the conventional bone window setting. In addition, it provided better soft-tissue visualization than the bone window. The software, by displaying clear outlines of screws and decreasing metal artifact, as well as by reconstructing the images in arbitrary planes, was more helpful in identifying the position of pedicle screws than the conventional bone window setting.