Radiological aspects of spine diagnostics and surgery

Background: Intraoperative imaging in spine surgery, such as lateral radiographs and 2D fluoroscopy, was first used in the 1970s. Cone beam computed tomography (CBCT), which utilizes 3D technology, was introduced in the 1990s. Initially developed for dental imaging, CBCT has expanded into various fields, including spine surgery, interventional procedures and musculoskeletal radiology. In spine surgery CBCT helps spine surgeons to identify and correct malplaced pedicle screws before finalizing the procedure, potentially eliminating revision surgeries. However, this imaging technology raises radiation exposure concerns for patients and staff. Furthermore, while postoperative multidetector computed tomography (MDCT) scans remain standard for confirming results, the convergence in image quality between CBCT and MDCT creates opportunities to streamline workflows and reduce redundant imaging, decreasing radiation exposure and healthcare costs. Among individuals with cervical spine injuries requiring surgical fixation, traumatic vertebral artery injury (VAI) presents a significant and potentially dangerous complication. While detection of VAI has improved through routine CT angiography (CTA), modern hybrid operating rooms (OR) equipped with CBCT technology used both in spine surgery and interventional procedures, offer a promising advancement: the ability to diagnose and treat both spinal and vascular injuries in a single session. This integrated approach could streamline patient care and potentially lead to better clinical outcomes.

Purpose: To investigate multiple aspects of spine surgery conducted within a hybrid OR. We evaluated occupational radiation exposure from various imaging equipment and different radiation protection shields (Paper I), compared the accuracy of CBCT with postoperative MDCT for detecting pedicle screw breaches (Paper II), assessed whether intraoperative CBCT provides equivalent image quality to postoperative MDCT to potentially reduce cumulative radiation exposure (Paper III), and examined the risk factors, incidence, outcomes, and imaging indicators of VAI in patients treated surgically for subaxial cervical spine injuries in trauma (Paper IV).

Methods: In Paper I, scatter radiation was measured in a hybrid OR while imaging an anthropomorphic phantom with three systems: ceiling-mounted hybrid C-arm cone beam CT (hCBCT), mobile O-arm CBCT (oCBCT), and mobile 2D C-arm fluoroscopy. Measurements were taken at various room positions utilizing active personal dosimetry devices and an ionization chamber. Two different radiation protection shields were assessed. Paper II evaluated 260 pedicle screws in twenty spinal fixation patients using intraoperative CBCT and follow-up MDCT. Three neurosurgeons independently graded pedicle screw breach using the Gertzbein scale, with MDCT as reference standard. The diagnostic evaluation methods assessed sensitivity, specificity, and negative predictive value. In Paper III, twenty-seven spinal fixation cases were examined, comparing intraoperative CBCT with postoperative MDCT. Four neuroradiologists independently evaluated images using Likert scales, and visual grading analysis measured modality preferences. Observer agreement was quantified through ICC analysis while image quality was objectively measured via contrast- and signal-to-noise ratios (CNR, SNR). In Paper IV, traumatic subaxial injuries that were surgically treated during a twelve-year period were analyzed, with mortality and morbidity as primary outcomes. Propensity score matching was used to create comparable groups, survival analysis tracked outcomes over time, single- and multiple- variable analyses examined the clinical outcomes.

Main Results: Paper I found that OR personnel were exposed to approximately 22% lower radiation when procedures utilized the hCBCT equipment compared to operations conducted with the oCBCT system. The scattered radiation doses emitted from the hCBCT were observed to be 27% less than those from the oCBCT when measured at 200 cm from the phantom. A single rotation for image acquisition utilizing hCBCT corresponded to 12 minutes of C-arm fluoroscopy, whereas the oCBCT corresponded to 16 minutes. The scatter dose decreased by over 90% behind radiation protection shielding, although this protective effect diminished slightly when measured at greater distances (60 cm), likely attributable to secondary radiation reflection from ceilings and walls. In Paper II intraoperative CBCT demonstrated a high negative predictive value of 99.6% in ruling out pedicle screw breaches. It also showed a sensitivity of 90.0% and a specificity of 97.6% in detecting screw accuracy compared to postoperative MDCT. In Paper III intraoperative CBCT was the superior modality in the thoracolumbar spine. Conversely, postoperative MDCT was preferred in cervical spine. The agreement was good for inter-observers and moderate in intra- observers. SNR and CNR were comparable in thoracolumbar imaging, while MDCT provided superior and more consistent image quality in the cervical spine. In Paper IV, our analysis revealed that VAI was predominantly associated with significant high-energy trauma, particularly collisions from motor vehicles, same-and high-level falls, with a demographic profile of male predominance with an age distribution centered at 64.4 years. Two-thirds of the cases with VAI also exhibited spinal cord trauma, with this injury combination linked to higher levels of neurological dysfunction. When adjusting for demographic variables our analysis revealed that VAI showed no significant impact on complications arising after surgery, immediate or extended recovery outcomes, or survival rates. Facet joint dislocation was a distinctive imaging indicator that predicted VAI.

Conclusions: In Paper I, hCBCT reduced occupational radiation exposure compared to oCBCT in image-guided spine surgery. Staff can lower radiation exposure through optimal positioning relative to the radiation source, patient, walls and ceilings. This allows for the use of RPSs instead of lead aprons during imaging, improving comfort while reducing whole-body dose. Intraoperative CBCT imaging in Paper II proved to be reliable and equivalent to postoperative MDCT scanning in detecting malpositioned pedicle screws in the thoracolumbar spine, potentially eliminating the need for an additional postoperative MDCT examination. In Paper III, CBCT offered superior image quality for thoracolumbar imaging, while MDCT was better suited for cervical imaging. Intraoperative CBCT could potentially replace postoperative MDCT for thoracolumbar spine procedures, though postoperative MDCT remains essential for cervical spine evaluation. Paper IV found that VAI did not negatively affect the clinical results in surgically treated cervical spine injuries. Although multiple imaging features showed associations with VAI, facet joint dislocation was the only standalone predictor of VAI.

List of scientific papers

I. Radiation distribution in a hybrid operating room, utilizing different X-ray imaging systems: investigations to minimize occupational exposure. Cewe P, Vorbau R, Omar A, Elmi-Terander A, Edström E. J Neurointerv Surg. 2022 Nov;14(11):1139-1144. https://doi.org/10.1136/neurintsurg-2021-018220

II. Intraoperative cone beam computed tomography is as reliable as conventional computed tomography for identification of pedicle screw breach in thoracolumbar spine surgery. Burström G*, Cewe P*, Charalampidis A, Nachabe R, Söderman M, Gerdhem P, Elmi-Terander A, Edström E. Eur Radiol. 2021 Apr;31(4):2349-2356. https://doi.org/10.1007/s00330-020-07315-5

III. Image Quality Assessment in Spine Surgery: A Comparison of Intraoperative CBCT and Postoperative MDCT. Cewe P, Skorpil M, Fletcher-Sandersjöö A, VG El-Hajj, Grane P, Fagerlund M, Kaijser M, Elmi-Terander A, Edström E. Acta Neurochir. 167, 94 (2025). https://doi.org/10.1007/s00701-025-06503-w

IV. Traumatic Vertebral Artery Injury After Subaxial Cervical Spine Injuries: Incidence, Risk Factors, and Long-Term Outcomes: A Population-Based Cohort Study. El-Hajj VG*, Habashy KJ*, Cewe P*, Atallah E, Singh A, Fletcher-Sandersjöö A, Bydon M, Fagerlund M, Jabbour P, Gerdhem P, Elmi-Terander A, Edström E. Neurosurgery. 2024 Sep 20. https://doi.org/10.1227/neu.0000000000003173