Prospective comparison of virtual fluoroscopy to fluoroscopy and plain radiographs for placement of lumbar pedicle screws.

Fluoroscopy-based frameless stereotactic systems provide feedback to the surgeon using virtual fluoroscopic images. The real-life accuracy of these virtual images has not been compared with traditional fluoroscopy in a clinical setting. We prospectively studied 23 consecutive cases. In two cases, registration errors precluded the use of virtual fluoroscopy. Pedicle probes placed with virtual fluoroscopic imaging were imaged with traditional fluoroscopy in the remaining 21 cases. Position of the probes was judged to be ideal, acceptable but not ideal, or not acceptable based on the traditional fluoroscopic images. Virtual fluoroscopy was used to place probes in for 97 pedicles from L1 to the sacrum. Eighty-eight probes were judged to be in ideal position, eight were judged to be acceptable but not ideal, and one probe was judged to be in an unacceptable position. This probe was angled toward an adjacent disc space. Therefore, 96 of 97 probes placed using virtual fluoroscopy were found to be in an acceptable position. The positive predictive value for acceptable screw placement with virtual fluoroscopy compared with traditional fluoroscopy was 99%. A probe placed with virtual fluoroscopic guidance will be judged to be in an acceptable position when imaged with traditional fluoroscopy 99% of the time.