Abstract
Background Minimally invasive transfacet transforaminal lumbar interbody fusion (TF-TLIF) offers advantages over open approaches, including reduced tissue disruption and faster recovery. However, limited visualization increases the risk of neural injury, particularly to the exiting nerve roots and thecal sac.
Innovation This case report illustrates the integration of augmented reality (AR) to enhance real-time visualization during TF-TLIF. A 53-year-old man with symptomatic L4 to L5 spondylolisthesis underwent AR-assisted TF-TLIF. Preoperative imaging included magnetic resonance imaging, computed tomography, and advanced neurography sequences (Multi-Echo iN Steady-state Acquisition, Short Tau Inversion Recovery, and Dixon T1), which were used to generate a 3D model of critical anatomy. A safe transfacet trajectory was planned preoperatively and overlaid onto the surgical field through the operative microscope, using intraoperative 3D imaging for registration. Intraoperative neurophysiological monitoring complemented visual guidance.
Clinical Relevance AR enabled continuous visualization of neural structures during drilling, discectomy, and cage placement. The patient had no complications and was discharged on postoperative day 1 without new neurological deficits. While limited to a single case, this report demonstrates the feasibility of AR-assisted TF-TLIF. This technology may serve as a promising adjunct in minimally invasive spine surgery. Further studies are needed to assess the impact on efficiency and outcomes.
Level of Evidence 4.
Introduction
Minimally invasive spine surgery has significantly advanced over the past few decades, providing a transformative alternative to traditional open spinal procedures. Among these techniques, minimally invasive transfacet transforaminal lumbar interbody fusion (TF-TLIF) has emerged as a method for treating a range of spinal pathologies, including degenerative disc disease, spondylolisthesis, and lumbar instability. By minimizing tissue disruption, TF-TLIF reduces surgical morbidity, shortens recovery times, and enhances postoperative outcomes.1,2 Compared with its open counterpart, TF-TLIF achieves these benefits by utilizing smaller incisions, reducing blood loss, and preserving muscular integrity.3,4 However, the reduced surgical exposure associated with minimally invasive techniques introduces unique challenges, particularly in achieving accurate visualization of critical neural structures such as the exiting nerve roots and the dural sac.5,6
Neural injury remains 1 of the most concerning complications in TF-TLIF, with the limited field of view increasing the risk of pedicle breach, misplaced instrumentation, and dural tears.7,8 Augmented reality (AR) technology has emerged as a promising solution to address these challenges by providing surgeons with enhanced real-time visualization of hidden anatomical structures during surgery. AR can be used to overlay critical 3D anatomical information onto the surgical field, allowing surgeons to navigate more precisely while mitigating risks.9,10 By integrating preoperative imaging data and intraoperative navigation systems, AR offers a level of clarity and accuracy that was previously unattainable with traditional imaging modalities.11,12 These benefits can be multiplied through the use of novel magnetic resonance imaging (MRI) sequences for segmentation. These include sequences commonly used for peripheral neurography, such as Multi-Echo iN Steady-state Acquisition (MENSA) and Short Tau Inversion Recovery (STIR) sequences.13,14
This technical report explores the integration of AR in a TF-TLIF procedure for the treatment of L4 to L5 spondylolisthesis (Figure 1). We outline an approach to combine preoperative 3D segmentation, real-time AR visualization, and intraoperative guidance. The objective of this work is to establish a feasible protocol for applying AR in the world of minimally invasive spine surgery.
Preoperative imaging, including (A) lateral flexion radiograph, (B) lateral extension radiograph, and (C) sagittal and (D) axial views on preoperative T2 imaging, showing significant stenosis at the L4/L5 disk.
Technique
We describe the integration of AR into minimally invasive TF-TLIF to enhance intraoperative visualization of neural structures. This approach combines high-resolution imaging, 3D modeling, and intraoperative navigation to guide the surgeon through a safe corridor to the disc space while minimizing risk to critical anatomy.
Preoperative Imaging and Segmentation
The workflow begins with the acquisition of several high-resolution MRI sequences. MENSA MRI was developed to image peripheral nerves and is used here to visualize distal branches of the lumbar plexus as they enter the surrounding muscle and connective tissue.15,16 A Dixon T1 MRI is employed to separate fat and water signals, enhancing the visualization of structures with varying tissue compositions.17 A STIR MRI is used to suppress fat signals.18 All of these are registered to a Cube T2 MRI in BrainLab software (BrainLab MedTech, Munich, Germany; Figure 2).
Coronal cross sections showing (A) Short Time Inversion Recovery, (B) Cube T2, (C) Multi-Echo iN Steady-state Acquisition (MENSA), and (D) T1 Dixon magnetic resonance images.
Tissue Segmentation, 3-Dimensional Modeling, and Preoperative Planning
Segmentation is performed in BrainLab following a consistent sequence (Figure 3A–C). First, each vertebra is roughly segmented with the smart brush on the baseline T2 MRI. The smart brush uses local thresholding in 3 dimensions to fill similar adjacent voxels, allowing for rapid volume generation. After this, the intervertebral disks are segmented, using the start and end of each vertebra as a guide. Similarly, the psoas muscles on each side of the vertebrae are roughly segmented. BrainLab’s automatic cerebrospinal fluid (CSF) segmentation tool is implemented to approximate the thecal sac.
Segmentation as viewed (A) from the back with all structures included with the safe trajectory represented by a screw, (B) from the same posterior view without the L2–L5 vertebrae and intervertebral disks illustrating the exit of the L2–L5 rootlets and their entry into the bilateral iliopsoas (blue), and finally (C) from the right side down the axis of the trajectory. Three views are also provided of the trajectory into the L4/L5 intervertebral disk: (D) axial, (E) sagittal, and (F) down the axis of the trajectory.
Nerve rootlets are segmented in descending order, allowing the technician to follow the natural course of each rootlet as it exits the thecal sac. Furthermore, this segmentation is performed voxel-by-voxel without the smart-brush tool, relying on the user’s anatomic knowledge rather than blind thresholding of surrounding voxels. This also allows for rapid use of MENSA, STIR, and Dixon T1 sequences side-by-side, helping the user follow each rootlet and branch through the thecal sac and into adjacent tissue.
Once rootlets have been followed out of the thecal sac, the user returns to clean and sculpt the vertebrae, intervertebral disks, and psoas. The foramina of each vertebra are easily identified by tracing the path of the exiting rootlet. In a similar fashion, the edges of bulging intervertebral disks are also more clearly delineated using the thecal sac and exiting rootlets. Finally, facet joints between adjacent vertebrae, as well as posterior laminae, are identified and cleaned in descending order. As with rootlets, this allows the user to apply anatomic knowledge regarding the spatial relationship between vertebrae.
Once segmentation is complete, a trans-facet trajectory is identified to the requisite disk space, avoiding branches of the lumbar plexus (Figure 3D–F). The minimum distance is also identified between the planned trajectory and critical tissues. A step-by-step approach is supplied in Supplementary Methods.
Intraoperative AR Setup and Application
A reference marker is placed on the iliac crest to enable accurate tracking intraoperatively. Three-dimensional images are then acquired using a mobile C-arm—in this case, the OEC 3D system (GE Healthcare, Chicago, IL). The neuronavigation system is then calibrated to align the surgical instruments with the patient’s anatomy precisely (Figure 4A–C). The preoperative MRIs and volumes of all segmented anatomy are subsequently warped into intraoperative computed tomography (CT) space using BrainLab Spine Curvature Correction Software (version 1, Figure 4D).
This illustration shows the verification procedure used to co-register the intraoperative computed tomography (CT) image with the preoperative magnetic resonance images and associated segmentation, starting with (A–C) co-localization of a stylette with metal landmarks, and ending with (D) an overlay of the intraoperative CT with the original segmentation. These allow for the operating surgeon to see (E) overlayed L4 (green) and L5 (light pink) rootlets on the surgical field as viewed through the operative microscope.
Overlay accuracy is assessed through the identification of clear landmarks through visual assessment and touch. Note that, as there was no nonlinear warp applied, registration errors would manifest as shifts of multiple anatomic landmarks at once, allowing for rapid feedback and correction. Anatomic landmarks and elements can subsequently be visualized intraoperatively as overlays through the surgical microscope (Figure 4E).
Clinical Application
Patient Selection
A 53-year-old man presented to our institution with low back pain and radiating sciatic right leg pain in the setting of L4 to L5 spondylolisthesis. His pain began 1 year prior to his initial presentation while he was performing physical labor, and it worsened with time. He described his pain as radiating down his right lateral thigh and calf, sometimes spreading along the plantar aspect of his foot. It was exacerbated by activity and extended periods of remaining seated in 1 position. At presentation, he denied any weakness, balance difficulties, or difficulties with bowel or bladder control. While he had previously tried various medications as well as physical therapy, his pain continued to worsen up to his presentation. Due to our patient’s worsening pain with radiologic evidence of L4 to L5 grade 1 anterolisthesis, we offered him an L4/L5 TF-TLIF (Figure 1). Informed consent was obtained prior to the procedure, and preoperative imaging and segmentation were completed as previously mentioned.
Surgical Approach
Patient Positioning
Under general anesthesia, the patient was positioned prone on a radiolucent operating table. Intraoperative neurophysiological monitoring (IONM) was utilized to provide real-time feedback on the functional integrity of the nerve roots during the procedure.
Percutaneous Pedicle Screw Insertion
Using pseudo-live instrument tracking fluoroscopy-guided navigation (TrackX, Hillsborough, NC), percutaneous pedicle screws were inserted at L4 and L5 via a bilateral Wiltse approach.19,20 The navigation system ensured accurate placement of the screws with minimal radiation exposure.21,22
AR Setup
A reference marker was placed on the left iliac crest in this patient. As previously mentioned, an intraoperative CT was obtained, and preoperative imaging was registered to this scan. The accuracy of this warp was assessed by tracing the edge of bony landmarks.
Access to the Surgical Site
Navigated dilators were used to access the right L4 to L5 facet. An expandable (Globus Medical, Audubon, PA) retractor was positioned with blades placed cranially and caudally, allowing free mediolateral manipulation of instruments. Lateral fluoroscopy confirmed accurate retractor positioning, at which point the retractors were secured with an articulated arm.
AR-Guided Transfacet Corridor Drilling
AR technology projected images of the L4 to L5 disc, the right L4 and L5 exiting nerve roots, and the dural sac onto the surgical microscope, providing detailed visualization for the surgeon (Figure 4E). After meticulous soft tissue removal, drilling was performed between the L4 and L5 facets, following the preplanned safe corridor to the disc. This involved drilling through the L4 inferior facet, the synovial interface, and the L5 superior facet. The real-time AR projection ensured drilling remained within the safe zone, continuously displaying the positions of critical neural structures. IONM was employed during this step to monitor the functional integrity of the nerve roots, complementing the visual guidance from AR.
Discectomy and Cage Placement
A discectomy was performed using pituitary rongeurs. An expandable cage, Dual-X (Amplify Surgical, Irvine, CA, USA), was carefully placed into the disc space to maintain proper spinal alignment and stability. Anteroposterior and lateral fluoroscopy confirmed precise cage placement (Figure 5).
Postoperative radiographs illustrating proper cage placement at L4/L5 from (A) coronal and (B) sagittal views.
Final Fixation and Closure
The Globus retractor and navigation reference were removed. Appropriate rods were positioned and secured to stabilize the spine. The surgical site was irrigated, and the incision was closed in standard fashion. The surgery lasted 3 hours and 50 minutes in total, with an estimated blood loss of 100 mL.
Postoperative Course
The patient tolerated the procedure well with no intraoperative complications. The use of AR provided enhanced visualization of neural structures, allowing for precise instrument placement. Postoperative imaging confirmed the correct placement of hardware and restoration of spinal alignment. The patient reported significant improvement in symptoms with no new neurological deficits. He was discharged home and walked out of the hospital on postoperative day 1.
Discussion
The integration of AR technology in minimally invasive spine surgery, specifically for procedures such as TF-TLIF, represents a promising technological advancement. This technical report outlines the practical application of AR in 1 such case, offering insight into its feasibility and workflow integration.
An advantage of AR in spine surgery is its ability to help the surgeon visualize hidden structures in real time. In traditional approaches to spine surgery, surgeons utilize retraction and anatomic dissection to identify landmarks and ensure safety. However, excessive retraction and dissection have been associated with iatrogenic complications such as radiculopathy, sensory loss, and motor weakness.23 By allowing surgeons to visualize these structures without dissection, AR has the potential to reduce extensive retraction, potentially reducing these complications. This is even more salient in minimally invasive approaches, where additional dissection may be more challenging.
Instrument accuracy is paramount in minimally invasive spine surgery, and prior work has demonstrated the value of projecting drilling trajectories through intraoperative displays.24 Our approach builds upon that foundation by overlaying not just bony anatomy but also critical soft tissue structures such as exiting nerve rootlets and the thecal sac. In this case, we found that this additional layer of visualization complemented existing navigation. Furthermore, the integration of IONM added functional assurance to the visual information provided by AR.25 Used together, these systems provide both anatomic and electrophysiological confirmation, which may be particularly useful in cases with altered anatomy, revision surgery, or limited exposure.
Although the AR-assisted workflow described here successfully complemented our minimally invasive approach, we acknowledge the increased time and preparation required. Operative duration in this case was longer than a standard minimally invasive TF-TLIF, and preoperative segmentation took time and anatomic knowledge.26 Future cases may benefit from workflow optimization and improved automation of preoperative planning. Longitudinal studies will be necessary to determine whether AR integration eventually improves workflow efficiency or outcomes.
AR technology has been increasingly adopted across multiple neurosurgical and orthopedic domains beyond the spine. In cranial neurosurgery, AR has been used for tumor localization, trajectory planning, and intraoperative guidance during endoscopic and open resections.27 In orthopedic surgery, AR has shown utility in joint arthroplasty, fracture fixation, and percutaneous instrumentation, particularly for enhancing precision and reducing fluoroscopy exposure.28 Broader context from these fields reinforces AR’s emerging role as a platform technology with cross-specialty relevance.
The primary limitation of this work is its single-case design, which prevents any conclusions about generalizability, safety, efficacy, or efficiency. Although this patient had a favorable outcome, larger studies across diverse pathologies and patient populations are needed to validate AR-assisted workflows. Additional barriers include the need for specialized hardware, software licensing, and trained personnel. These aspects may severely limit adoption in resource-constrained settings. Furthermore, the segmentation and registration processes remain semimanual and time consuming, although future automation may mitigate this barrier.29 Finally, understanding the learning curve for AR implementation and the cost-effectiveness of these tools will be critical to guide broader implementation and protocol standardization.30
Conclusions
The application of AR in minimally invasive spine surgery, particularly in TF-TLIF, represents a promising step forward in surgical innovation. This technical report demonstrates the feasibility of using AR to enhance intraoperative visualization of neural structures and guide surgical instrument placement with improved anatomical context. When paired with IONM, AR may offer an additional layer of precision and safety, particularly in anatomically constrained or high-risk cases. Future work is required to evaluate the impact of AR on safety and outcomes, especially among different institutions. Additional work is also required to optimize the integration of AR platforms and automated segmentation tools into the surgical workflow. With continued refinement, AR has the potential to improve accuracy and decision-making in spine surgery, elevating the standard of care in minimally invasive techniques.
Acknowledgments
We acknowledge and thank the operating room staff who make these procedures possible for surgical teams and patients.
Footnotes
Funding The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests The authors report no conflicts of interest in this work.
Disclosures Dr. Abd-El-Barr is a consultant for BrainLab, Globus Medical, TrackX, and Arthrex.
Ethics Statement Institutional review board approval was not required for this case report, and patient consent was acquired.
- This manuscript is generously published free of charge by ISASS, the International Society for the Advancement of Spine Surgery. Copyright © 2025 ISASS. To see more or order reprints or permissions, see http://ijssurgery.com.
References
- 1.↵Wong AP , Smith ZA , Stadler JA , et al . Minimally invasive transforaminal lumbar interbody fusion (MI-TLIF): surgical technique, long-term 4-year prospective outcomes, and complications compared with an open TLIF cohort. Neurosurg Clin N Am. 2014;25(2):279–304.
- 2.↵Lee W-C , Park J-Y , Kim KH , et al . Minimally invasive transforaminal lumbar interbody fusion in multilevel: comparison with conventional transforaminal interbody fusion. World Neurosurg. 2016;85:236–243. 10.1016/j.wneu.2015.09.009
- 3.↵Mobbs RJ , Sivabalan P , Li J . Technique, challenges and indications for percutaneous pedicle screw fixation. J Clin Neurosci. 2011;18(6):741–749. 10.1016/j.jocn.2010.09.019
- 4.↵Härtl R , Lam KS , Wang J , Korge A , Kandziora F , Audigé L . Worldwide survey on the use of navigation in spine surgery. World Neurosurg. 2013;79(1):162–172. 10.1016/j.wneu.2012.03.011
- 5.↵Contreras López WO , Navarro PA , Crispin S . Intraoperative clinical application of augmented reality in neurosurgery: a systematic review. Clin Neurol Neurosurg. 2019;177:6–11. 10.1016/j.clineuro.2018.11.018
- 6.↵Elmi-Terander A , Skulason H , Söderman M , et al . Surgical navigation technology based on augmented reality and integrated 3D intraoperative imaging: a spine cadaveric feasibility and accuracy study. Spine (Phila Pa 1986). 1976;41(21):E1303–E1311. 10.1097/BRS.0000000000001830
- 7.↵Virk S , Qureshi S . Navigation in minimally invasive spine surgery. J Spine Surg. 2019;5(Suppl 1):S25–S30. 10.21037/jss.2019.04.23
- 8.↵Molina CA , Theodore N , Ahmed AK , et al . Augmented reality-assisted pedicle screw insertion: a cadaveric proof-of-concept study. J Neurosurg Spine. 2019;31(1):139–146. 10.3171/2018.12.SPINE181142
- 9.↵Gibby JT , Swenson SA , Cvetko S , Rao R , Javan R . Head-mounted display augmented reality to guide pedicle screw placement utilizing computed tomography. Int J Comput Assist Radiol Surg. 2019;14(3):525–535. 10.1007/s11548-018-1814-7
- 10.↵Keating TC , Jacobs JJ . Augmented reality in orthopedic practice and education. Orthop Clin North Am. 2021;52(1):15–26. 10.1016/j.ocl.2020.08.002
- 11.↵Dennler C , Jaberg L , Spirig J , et al . Augmented reality-based navigation increases precision of pedicle screw insertion. J Orthop Surg Res. 2020;15(1):174. 10.1186/s13018-020-01690-x
- 12.↵Elmi-Terander A , Burström G , Nachabe R , et al . Pedicle screw placement using augmented reality surgical navigation with intraoperative 3D imaging. Spine (Phila Pa 1986). 2019;44(7):517–525. 10.1097/BRS.0000000000002876
- 13.↵Pan X , Wen Y , Huang K , et al . Application of three-dimensional (3D) magnetic resonance (MR) multi-echo in steady-state acquisition sequences in preoperative evaluation of lumbar disc herniation: a prospective study. Quant Imaging Med Surg. 2024;14(10):7540–7550. 10.21037/qims-23-1834
- 14.↵Kristoffersen PM , Vetti N , Storheim K , et al . Short tau inversion recovery MRI of modic changes: a reliability study. Acta Radiol Open. 2020;9(1):2058460120902402. 10.1177/2058460120902402
- 15.↵Eckstein F , Hudelmaier M , Wirth W , et al . Double echo steady state magnetic resonance imaging of knee articular cartilage at 3 tesla: a pilot study for the osteoarthritis initiative. Ann Rheum Dis. 2006;65(4):433–441. 10.1136/ard.2005.039370
- 16.↵Hardy PA , Recht MP , Piraino D , Thomasson D . Optimization of a dual echo in the steady state (DESS) free-precession sequence for imaging cartilage. J Magn Reson Imaging. 1996;6(2):329–335. 10.1002/jmri.1880060212
- 17.↵Lins CF , Salmon CEG , Nogueira-Barbosa MH . Applications of the dixon technique in the evaluation of the musculoskeletal system. Radiol Bras. 2021;54(1):33–42. 10.1590/0100-3984.2019.0086
- 18.↵Golfieri R , Baddeley H , Pringle JS , Souhami R . The role of the STIR sequence in magnetic resonance imaging examination of bone tumours. Br J Radiol. 1990;63(748):251–256. 10.1259/0007-1285-63-748-251
- 19.↵Wiltse LL . The paraspinal sacrospinalis-splitting approach to the lumbar spine. Clin Orthop Relat Res. 1973;91(91):48–57. 10.1097/00003086-197303000-00009
- 20.↵Guiroy A , Sícoli A , Masanés NG , Ciancio AM , Gagliardi M , Falavigna A . How to perform the wiltse posterolateral spinal approach: technical note. Surg Neurol Int. 2018;9:38. 10.4103/sni.sni_344_17
- 21.↵Wang TY , Tabarestani TQ , Mehta VA , et al . A comparison of percutaneous pedicle screw accuracy between robotic navigation and novel fluoroscopy-based instrument tracking for patients undergoing instrumented thoracolumbar surgery. World Neurosurg. 2023;172:e389–e395. 10.1016/j.wneu.2023.01.037
- 22.↵Hamouda F , Wang TY , Gabr M , et al . A prospective comparison of the effects of instrument tracking on time and radiation during minimally invasive lumbar interbody fusion. World Neurosurg. 2021;152:e101–e111. 10.1016/j.wneu.2021.05.058
- 23.↵Gerling MC , Baker M , Stanton E , Chaladoff E , Buser Z . Nerve root retraction time during lumbar endoscopic discectomy: association with new onset radiculitis, a post-operative neurologic complication. Eur Spine J. 2024;33(1):126–132. 10.1007/s00586-023-07952-6
- 24.↵Bardeesi A , Tabarestani TQ , Bergin SM , et al . Using augmented reality technology to optimize transfacet lumbar interbody fusion: a case report. J Clin Med. 2024;13(5):1513. 10.3390/jcm13051513
- 25.↵Gonzalez AA , Jeyanandarajan D , Hansen C , Zada G , Hsieh PC . Intraoperative neurophysiological monitoring during spine surgery: a review. Neurosurg Focus. 2009;27(4):E6. 10.3171/2009.8.FOCUS09150
- 26.↵Zhang X , Zhang Y , Gu Z , Li G . Comparison of midline lumbar interbody fusion and minimally invasive transforaminal lumbar interbody fusion for treatment of lumbar degeneration disease. Sci Rep. 2024;14(1):22154. 10.1038/s41598-024-73213-1
- 27.↵Van Gestel F , Frantz T , Buyck F , et al . Neuro-oncological augmented reality planning for intracranial tumor resection. Front Neurol. 2023;14:1104571. 10.3389/fneur.2023.1104571
- 28.↵Rossi SMP , Mancino F , Sangaletti R , Perticarini L , Lucenti L , Benazzo F . Augmented reality in orthopedic surgery and its application in total joint arthroplasty: a systematic review. Appl Sci (Basel). 2022;12(10):5278. 10.3390/app12105278
- 29.↵Niemeyer F , Zanker A , Jonas R , Tao Y , Galbusera F , Wilke HJ . An externally validated deep learning model for the accurate segmentation of the lumbar paravertebral muscles. Eur Spine J. 2022;31(8):2156–2164. 10.1007/s00586-022-07320-w
- 30.↵Bischofreiter M , Gabauer N , Gruber MS , et al . Learning curve of augmented reality surgical application for reverse shoulder arthroplasty. Arch Orthop Trauma Surg. 2025;145(1):252. 10.1007/s00402-025-05862-4
