Abstract
Background Fluoroscopy and computed tomography (CT) are standard imaging modalities for spinal interventions but expose patients and health care professionals to ionizing radiation. Ultrasonography–CT (US–CT) fusion systems enable co-registration of real-time ultrasonography with pre-existing CT images, offering enhanced anatomical visualization without radiation. This study evaluates the feasibility and accuracy of a US–CT fusion system for vertebral level identification.
Materials and Methods Seventeen patients scheduled for lumbar CT were enrolled. Using a GE Logiq E9 US–CT fusion system, lumbar vertebrae (L1–L5) underwent imaging in the prone position, with facet joints and sacroiliac joints as coregistration landmarks. A randomly selected vertebra was marked with a pin, and its level was verified by a subsequent CT image.
Results In 17 patients, 15 vertebral levels (88%) were correctly identified using the US–CT fusion system. Facet joints and sacroiliac joints were reliable landmarks for coregistration, enabling synchronized ultrasonography and CT visualization. Two errors occurred, likely due to pin displacement or misidentification.
Conclusions The US–CT fusion system is a promising tool for lumbar vertebral identification, with the potential to reduce radiation exposure during spinal procedures. This observational study has several limitations, including a small sample size and an operator learning curve. Further studies are needed to optimize its application; however, its potential clinical relevance remains high.
Introduction
Spine disorders are extremely prevalent and are associated with significant medical morbidity and high health care costs.1–4 Among the available treatment options for various spinal pathologies are surgical procedures such as laminectomy and foraminotomy, as well as minimally invasive interventions such as facet joint injections and nerve root blocks.5,6 All of these treatments require imaging guidance to enable precise targeting of the anatomical region responsible for the patient’s clinical presentation. Minimally invasive injections and surgical procedures are most commonly performed under computed tomography (CT) or x-ray guidance, both of which involve exposure to ionizing radiation and typically require the presence of a trained technician.7,8
One of the commonly used imaging modalities in modern medicine is ultrasonography (US), which has clear advantages, including the ability to image soft tissues, its high availability, its relatively low cost, and the absence of radiation exposure.9 US has also been employed in the diagnosis and treatment of spine pathologies. For example, US has been used to image back muscles and evaluate spine range of motion.10,11 US-guided epidural injections or foraminal injections have also been reported to be effective; however, their application in spinal injections is currently considered very limited.12 Despite the high availability of US, in recent years, its use has declined due to poor visualization of bony structures, and it is now primarily employed for the evaluation of spinal pathologies in neonates.13
Recently, US–CT fusion systems were introduced.14,15 These systems rely on the coregistration of a US device with an existing CT image. Several anatomical landmarks are used to perform coregistration, similar to intraoperative navigation systems.16 This creates a dual-screen interface, where the operator uses a US probe to visualize the anatomy in real time, while simultaneously viewing the corresponding CT image, allowing for improved anatomical visualization. These types of systems have recently been tested in several studies. For example, we previously reported their use in detecting and localizing urethral stones.17,18 In the current study, we aimed to evaluate this system for spinal imaging—specifically, whether a fusion system can identify different vertebral levels and what level of accuracy would make it feasible for use in marking and performing invasive procedures in the spine.
Methods
This was a single-center study conducted at a level 1 referral medical center with more than 900 beds. Ethical approval to conduct the study was granted by the institutional ethics committee (Helsinki Committee). Patients who were already scheduled for elective lumbar spine CT and had previously CT images performed at our institution were enrolled in the study. Exclusion criteria included any medical condition that would prevent lying in the prone position for several minutes, age younger than 18 years, or a prior surgical intervention involving the spine performed after the previous lumbar CT scan.
For all patients, we used a US–CT fusion system (GE Logiq E9 Ultrasound Machine). The system was operated by a senior certified radiologist. Initially, the patient’s previous CT scan was uploaded onto the US machine. The setup hardware and software for fusion were selected and used. Fusion requires using an electromagnetic transmitter positioned close to the patient and electromagnetic sensors on the US probe (a curved 1–5 MHz transducer). The patient was positioned prone next to the fusion system (Figure 1). Subsequently, US images were captured using the described transducer. Anatomical structures that were clearly visible on both US and CT images were used as landmarks to allow coregistration of the live US images with the pre-existing CT image and were selected manually on both datasets.
Fusion system setup. The patient is positioned prone, and the lumbar spine is scanned with the ultrasound (US) probe. An electromagnetic transducer is placed close to the patient. Coregistration is performed between computed tomography (CT) and real-time ultrasonography to allow a dual window view on the monitor.
A standardized protocol was developed during the study to optimize the imaging process. First, the sacroiliac joint (SIJ) was identified to establish the initial reference point, after which the system was recalibrated. Navigation between different vertebral levels was facilitated by identifying the facet joints, spinous processes, and transverse processes, with the facet joints serving as the most reliable anatomical landmarks for coregistration based on their clear visualization in both imaging modalities. Once the 2 imaging modalities were coregistered, the lumbar vertebrae underwent US from L5 to L1. The facet joints and the transverse and spinous processes were identified. Fused US–CT images in the same plane were moved synchronously during live US. The data from the CT image (Figure 1) were used as a guiding reference and enhanced the visualization of the anatomy. Once all lumbar vertebrae were scanned, a random lumbar vertebra was selected. The fusion system was utilized to determine the level of the selected vertebra (L5–L1), and a pin was placed on the patient’s skin to mark the corresponding vertebral height. The patient then proceeded to scheduled elective CT. Finally, we examined whether our assessment of vertebral level was correct by determining which vertebral level was marked by the pin in the subsequent CTimage.
Results
Seventeen patients underwent lumbar US using the fusion system and were included in the present study. Patient ages ranged from 29 to 88 years. As there are no published protocols for the use of US–CT fusion in the spine, we initially experimented to identify appropriate anatomical landmarks for coregistration. We found that the facet joints and the SIJ were reliable markers for establishing coregistration, as they were clearly visible on both imaging modalities. Once coregistration was completed, real-time US images were synchronized with the CT window, allowing improved anatomical visualization (Figure 2).
Dual window view. Once coregistration was completed, live ultrasound images were synchronized to the computed tomography data, allowing visualization of the anatomy in two separate imaging modalities. (A) Axial view of the lumbar spine. (B) Axial view of the sacrum. (C) Sagittal view of the lumbar spine.
Among the 17 patients who underwent US, we selected the suspected L3 vertebra for marking in 14 patients, the suspected L4 vertebra in 1 patient, and the L2/L3 intervertebral disc in 2 patients (Table). The correct vertebral level, confirmed by the pin location in the subsequent CT image (Figure 3), was identified in 15 patients (88%; 95% CI: 64%–98%). In 1 incorrect case, L4 was misidentified as L3. In another case, the location of the pin deviated significantly from the expected height, appearing at the level of S1 instead of L3. We suspect that this significant error resulted from the detachment of the pin sticker prior to the CT scan. However, skin-to-bone migration during patient transfer from the prone to the supine position may also account for the vertebral level misidentification.
Confirmation of assigned vertebra. The computed tomography scan performed at the end of the ultrasound scan is reviewed, and the level of the assigned vertebra is confirmed. (A) Sagittal section. (B) Axial section. Red arrows mark the pasted pin used to mark the assigned vertebra.
Assigned and marked vertebra in each patient.
Discussion
The fusion system is an innovative technology that enables real-time coregistration between US and other imaging modalities such as CT and MRI. By integrating these modalities, the system allows for simultaneous visualization and comparison between real-time US images and previously acquired images, which offer superior anatomical detail. This report is a pilot study to evaluate the application of this system in the spine, specifically its ability to localize and visualize individual vertebrae and relevant anatomical structures within the spinal column.
Previously, two studies have evaluated US–CT fusion on lumbar spine phantoms. Guntz et al used US–CT fusion to target the L4 to L5 foramina, effectively simulating a nerve root injection.19 Three different combinations of anatomical structures were used for coregistration and compared, and x-ray imaging was employed to confirm needle positioning. This study also demonstrated high accuracy of the fusion system in targeting specific spinal structures. Similarly, Chan et al reported comparable findings in a study conducted on thoracic spine phantoms.20 In this study, 3D US–CT fusion was utilized, and the mean image registration accuracy was reported as 0.3 mm, suggesting a feasible clinical application. While these studies are valuable in demonstrating the technical capabilities of the fusion system and offer a convenient setting for calibration, it remains difficult to assess the clinical accuracy and feasibility of US fusion technology without clinical trials.
In this prospective clinical study, we found that it is possible to identify the level of a vertebra with a relatively high level of accuracy (88%). This accuracy is comparable to that reported in our previous study on the use of this technology for identifying urethral stones.17 Furthermore, we developed a working protocol for applying this system to the spine by identifying the SIJ and facet joints as anatomical landmarks for coregistration. This approach allowed us to shift between different vertebral levels in real-time US while simultaneously observing the corresponding structures in the CT window.
One major limitation of this study is the small number of patients included. The wide confidence interval necessitates caution and highlights the need for larger trials. Furthermore, while an 88% identification rate may be considered high for an experimental early-stage system, this rate will need to be improved before the system can be considered feasible for clinical use. The accuracy of this system when applied to the spine may be influenced by variability in patient anatomy. However, we believe that with further refinement, the system could achieve higher accuracy. Therefore, additional studies are required not only to increase the number of participants but also to further refine the technology and technique before concluding that this technology is clinically applicable for spine procedures.
A number of modifications can be addressed specifically to increase the accuracy of this system. For example, we relied on skin markers, which are not secure enough and may shift with patient movement. Re-evaluating the vertebral level identification rate using more stable markers, such as needles, may help overcome the reliability limitations of skin markers, particularly if this technology is intended to guide needles for accurate injections. During the study, we also noticed that there was a learning curve associated with operating the fusion system effectively. We estimate that approximately 5 cases are required for a musculoskeletal radiologist to achieve proficient use of the system, suggesting that larger studies are required to obtain more robust results.
This study focused solely on the lumbar region of the spinal column; the thoracic and cervical regions were not scanned. As a result, further research is also needed to evaluate the system’s accuracy in these regions and find anatomical markers to calibrate the fusion system for the cervical and thoracic spine, in addition to refining scanning techniques in the lumbar spine to improve accuracy. Future studies could also focus on identifying specific vertebral structures, such as the lamina or pedicle, while assessing the system’s utility in the hands of spine surgeons.
Conclusion
The US–CT fusion system holds promise as a valuable tool for the management of spinal pathologies. It may serve as a supportive technology in surgical procedures that require precise identification of vertebral levels. Furthermore, it could assist in guiding injections for pain relief, such as nerve root blocks. Although additional research is warranted, we believe that implementing this technology has the potential to reduce dependence on radiation-based imaging while offering enhanced, real-time anatomical visualization.
Footnotes
Funding The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests The authors declare no conflicts of interest pertaining to this study.
Disclosures Generative artificial intelligence (ChatGPT 4) was used to improve spelling, grammar, and general editing. Figure 1 was prepared with generative artificial intelligence assistance. The authors retain full responsibility for the accuracy, interpretation, and integrity of the content presented.
- This manuscript is generously published free of charge by ISASS, the International Society for the Advancement of Spine Surgery. Copyright © 2026 ISASS. To see more or order reprints or permissions, see http://ijssurgery.com.
References
- 1.↵Katz JN , Zimmerman ZE , Mass H , Makhni MC . Diagnosis and management of lumbar spinal stenosis: a review. JAMA. 2022;327(17):1688–1699. 10.1001/jama.2022.5921
- 2.↵Friedly J , Standaert C , Chan L . Epidemiology of spine care: the back pain dilemma. Phys Med Rehabil Clin N Am. 2010;21(4):659–677. 10.1016/j.pmr.2010.08.002
- 3.↵Zhang C , Lv B , Yi Q , Qiu G , Wu F . Global, regional, and national burden of low back pain in working-age population from 1990 to 2021 and projections for 2050. Front Public Health. 2025;13:1559355. 10.3389/FPUBH.2025.1559355/BIBTEX
- 4.↵Menezes CM , Tucci C , Tamai K , et al . SPINE20 recommendations 2024 -spinal disability: social inclusion as a key to prevention and management. Global Spine J. 2025;15(1):8–20. 10.1177/21925682241290226
- 5.↵Phan K , Mobbs RJ . Minimally invasive versus open laminectomy for lumbar stenosis. Spine (Phila Pa 1986). 2016;41(2):E91–E100. 10.1097/BRS.0000000000001161
- 6.↵Chang KV , Wu WT , Özçakar L . Ultrasound-guided interventions of the cervical spine and nerves. Phys Med Rehabil Clin N Am. 2018;29(1):93–103. 10.1016/j.pmr.2017.08.008
- 7.↵Hadelsberg UP , Harel R . Hazards of ionizing radiation and its impact on spine surgery. World Neurosurg. 2016;92:353–359. 10.1016/j.wneu.2016.05.025
- 8.↵Baig AA , Anagnostou EN , Green K , Budu A . Radiation exposure and surgeon safety during minimally invasive spine surgery in the absence of neuronavigation: a UK single-centre retrospective study. Cureus. 2025;17(12):e99209. 10.7759/cureus.99209
- 9.↵Wagner LK , Eifel PJ , Geise RA . Potential biological effects following high x-ray dose interventional procedures. J Vasc Interv Radiol. 1994;5(1):71–84. 10.1016/s1051-0443(94)71456-1
- 10.↵Ahmed AS , Ramakrishnan R , Ramachandran V , Ramachandran SS , Phan K , Antonsen EL . Ultrasound diagnosis and therapeutic intervention in the spine. J Spine Surg. 2018;4(2):423–432. 10.21037/jss.2018.04.06
- 11.↵Chleboun GS , Amway MJ , Hill JG , Root KJ , Murray HC , Sergeev AV . Measurement of segmental lumbar spine flexion and extension using ultrasound imaging. J Orthop Sports Phys Ther. 2012;42(10):880–885. 10.2519/jospt.2012.3915
- 12.↵Chi M , Chen AS . Ultrasound for lumbar spinal procedures. Phys Med Rehabil Clin N Am. 2018;29(1):49–60. 10.1016/j.pmr.2017.08.005
- 13.↵Rohrschneider WK , Forsting M , Darge K , Tröger J . Diagnostic value of spinal US: comparative study with MR imaging in pediatric patients. Radiology. 1996;200(2):383–388. 10.1148/radiology.200.2.8685330
- 14.↵Ahn SJ , Lee JM , Lee DH , et al . Real-time US-CT/MR fusion imaging for percutaneous radiofrequency ablation of hepatocellular carcinoma. J Hepatol. 2017;66(2):347–354. 10.1016/j.jhep.2016.09.003
- 15.↵Sandulescu DL , Dumitrescu D , Rogoveanu I , Saftoiu A . Hybrid ultrasound imaging techniques (fusion imaging). World J Gastroenterol. 2011;17(1):49–52. 10.3748/wjg.v17.i1.49
- 16.↵O’Brien JR . The use of intraoperative CT and navigation for the treatment of spinal deformity in open and minimally invasive surgery. Spine (Phila Pa 1986). 2017;42(7):S28–S29. 10.1097/BRS.0000000000002038
- 17.↵Bercovich E , Leiderman M , Beck-Razi N , Gaitini D , Javitt M . Ultrasound–unenhanced CT fusion for detection and localization of ureteral stones. Am J Roentgenol. 2018;210(1):W8–W11. 10.2214/AJR.17.17840
- 18.↵Zisman A , Malshy K , Rinott B , et al . Ultrasound–CT fusion in ureteral stone follow-up: a prospective accuracy study benchmarked against ureteroscopy. Urolithiasis. 2025;53(1):203. 10.1007/S00240-025-01876-0/FIGURES/3
- 19.↵Emmanuel G , Arnaud P , Jonathan G , et al . Real-time ultrasound-computed tomography image fusion for transforaminal lumbar approach: a lumbosacral spine phantoms study. Eur Spine J. 2021;30(5):1270–1276. 10.1007/s00586-021-06794-4
- 20.↵Chan A , Coutts B , Parent E , Lou E . Development and evaluation of CT-to-3D ultrasound image registration algorithm in vertebral phantoms for spine surgery. Ann Biomed Eng. 2021;49(1):310–321. 10.1007/S10439-020-02546-5/FIGURES/8
