Warning: mkdir(): Permission denied in /home/virtual/lib/view_data.php on line 87 Warning: chmod() expects exactly 2 parameters, 3 given in /home/virtual/lib/view_data.php on line 88 Warning: fopen(/home/virtual/e-kjs/journal/upload/ip_log/ip_log_2024-12.txt): failed to open stream: No such file or directory in /home/virtual/lib/view_data.php on line 95 Warning: fwrite() expects parameter 1 to be resource, boolean given in /home/virtual/lib/view_data.php on line 96 A Self-Developed Mobility Augmented Reality System Versus Conventional X-rays for Spine Positioning in Intraspinal Tumor Surgery: A Case-Control Study
Neurospine Search

CLOSE


Neurospine > Volume 21(3); 2024 > Article
Hong, Huang, Li, Luo, Liu, Huang, Chen, He, Wen, and Lin: A Self-Developed Mobility Augmented Reality System Versus Conventional X-rays for Spine Positioning in Intraspinal Tumor Surgery: A Case-Control Study

Abstract

Objective

To evaluate the efficacy of a self-developed mobile augmented reality navigation system (MARNS) in guiding spinal level positioning during intraspinal tumor surgery based on a dual-error theory.

Methods

This retrospective study enrolled patients diagnosed with intraspinal tumors admitted to Fujian Provincial Hospital between May and November 2023. The participants were divided into conventional x-rays and self-developed MARNS groups according to the localization methods they received. Position time, length of intraoperative incision variation, and location accuracy were systematically compared.

Results

A total of 41 patients (19 males) with intraspinal tumors were included, and MARNS was applied to 21 patients. MARNS achieved successful lesion localization in all patients with an error of 0.38±0.12 cm. Compared to x-rays, MARNS significantly reduced positioning time (129.00±13.03 seconds vs. 365.00±60.43 seconds, p<0.001) and length of intraoperative incision variation (0.14 cm vs. 0.67 cm, p=0.009).

Conclusion

The self-developed MARNS, based on augmented reality technology for lesion visualization and perpendicular projection, offers a radiation-free complement to conventional x-rays.

INTRODUCTION

Intraspinal tumors arise from various components of the spinal structure, with an incidence ranging from 1.0 to 1.5 per 100,000 population [1]. While these tumors are often benign, surgical intervention is typically the primary treatment approach [2,3]. However, the suboptimal performance of spinal surgery at an incorrect level poses a significant safety risk [4].
Research indicates that wrong-level spine surgeries, where the surgical procedures including lesion exploration and resection is mistakenly performed at the wrong spinal level after opening the lamina, are more prevalent than previously thought, affecting up to 2% of patients [5,6], and resulting in heightened morbidity, increased costs, and poorer long-term outcomes [7]. In order to minimize this risk, surgeons commonly utilize intraoperative radiographs to confirm the correct surgical site [8]. However, the distinctive characteristics of spine anatomy continue to lead to incorrect-level procedures [9,10].
Furthermore, the frequent use of intraoperative x-rays causes an accumulation of ionizing radiation exposure [11]. Although surgical navigation systems currently available in the market are intended to improve surgical precision and reduce radiation exposure [12], the high cost, cumbersome nature, and time-consuming operation often impede their widespread implementation [13,14]. Other radiation-free localization methods, such as bony anatomical markers or methylene blue, often suffer from poor positioning accuracy. Therefore, we aspire to devise a radiationfree spinal level localization technique that is not only cost-effective and user-friendly but also guarantees precision, addressing the need for a practical and efficient alternative in spinal surgeries.
Augmented reality (AR) presents a viable solution to this challenge. By leveraging AR technology, surgeons can superimpose preoperative planning onto the intraoperative anatomy, enabling direct surgical navigation within their field of vision [15]. Empirical evidence suggests that AR-assisted spine surgery improves precision [16]. Given the exploration of mobile devices to display intraoperative AR views, constructing a mobile AR-based navigation system (MARNS) by utilizing consumer-grade materials such as tablets or mobile phones is a viable option at a substantially reduced cost compared to traditional surgical navigation systems [17]. Consequently, we proposed the implementation of a MARNS to facilitate surgical navigation for intraspinal tumors.
Anatomical variability and physician complacency may lead to wrong-level spine surgery [18,19]. However, the precise mechanism underlying this phenomenon remains uncertain. We postulated that the rates of incorrect-level spine surgery might be attributed to the confluence of 2 factors/errors: surgeons’ neglect of the physiological curvatures of the spine and their ingrained visual habits. Specifically, the physiological curvature of the spine may result in an inclined incision site in the prone position, and surgeons may be disregarded due to their customary practices, which is the first error. Consequently, when the skin is incised to expose the vertebral arch, the surgeon’s perspective is inherently orthogonal to the surgical site rather than orthogonal to the ground. The targeted spinal segment may stray from the center of the surgical field, while surgeons still consider it to be in the center, which is the second error (Fig. 1). This is what we refer to as our proposed double error theory on the occurrence of wrong-level spine surgery. In this study, based on the double error theory, we utilized a self-designed MARNS for tumor localization during surgical interventions.

MATERIALS AND METHODS

1. Study Design and Population

This retrospective study included patients diagnosed with intraspinal tumors admitted to Fujian Provincial Hospital between May 2023 and November 2023. The inclusion criteria were: (1) patients diagnosed with intraspinal space-occupying lesions by imaging and (2) scheduled for intraspinal tumor surgery. The exclusion criteria were: patients with adiposity, spinal deformity, and visible skin folds when in the prone position. The patients were divided into the conventional x-rays and MARNS groups according to the localization methods they received; patients who received conventional x-rays were included in the x-rays group, while those who received MARNS were assigned to the MARNS group. This study was conducted with the requisite ethics approval from the Ethics Committee of Fujian Provincial Hospital (No. K2020-06-019). All patients provided informed written consent. The study protocol was consistent with the principles of the 2013 Declaration of Helsinki and the International Conference on Harmonization Guidelines for Good Clinical Practice.

2. 3D Virtual Model Reconstruction

Two electrode patches were affixed to the patient’s back, followed by a thin-layer computed tomography (CT) scan (1 layer per 1 mm, no intervals) in the prone position. The CT scan data were imported into Mimics software (Version 21), and the dynamic adaptive region growing algorithm was employed to segment the target regions. Various thresholds were applied based on gray value differences between tissues. Considering the suspected double errors, the lesion was projected perpendicularly to the dorsal body surface during the 3-dimentional (3D) model reconstruction. Ultimately, the model was reconstructed by incorporating the body surface, vertebrae, electrode patches, and the tumor with its corresponding projection ticks (Fig. 2A). For cases when the tumor was not visible on a CT scan, a magnetic resonance image was matched with the CT image (Fig. 3). The method and process were showed in Supplementary Materials.

3. Development of MARNS

A mobile device (Vivo X60, Vivo, Dongguan, China) operating on the Android platform and an iron plate with a quick response (QR) code as fundamental hardware components were used in this study. The virtual 3D model and QR code image were imported in the Unity engine (ver. 2019.4.31, Unity Technologies, San Francisco, CA, USA). A specific metal point on the electrode patches within the 3D model was designated as the central reference point. Using this reference point, the QR code image was accurately positioned on the dorsal region of the virtual body surface. Subsequently, a spatial relationship was established between the 3D model and the QR code within the virtual reality environment of the Unity engine. Finally, an apapplication was developed for future deployment on an Android smartphone device (Fig. 4). The details regarding the development of MARNS application were showed in the Supplementary Materials.

4. MARNS Application and X-ray

We utilized a prone position for both CT scanning and surgical procedures to minimize potential variations and ensure consistency in imaging and surgical approach, thereby reducing discrepancies. During the surgical procedure, we carefully positioned a QR code iron plate at a predetermined location on the patient’s back, serving as a reference point for subsequent scanning. Using an Android smartphone equipped with our developed application, we scanned the QR code in both frontal and lateral orientations. This scanning process facilitated the display of the reconstructed 3D model (Fig. 2B and C) that accurately represented the patient’s anatomy. Specifically, we first marked the skin based on the tumor’s location indicated by the 3D model from the frontal view of the MARNS. To ensure the accuracy of this marking, we then conducted a lateral examination and employed the protractor measurement program (V2.12, developed in China) to confirm precise alignment with the vertical projection points of the intraspinal lesion onto the dorsal skin surface. This crucial alignment confirmation guaranteed the accuracy and effectiveness of the upcoming surgical intervention. If a substantial divergence was noticed between the 3D virtual model and the real-world scenario, we vertically projected the 3D tumor model presented by MARNS onto the patient’s actual dorsal body surface in the lateral view, thus pinpointing the incision center. A comprehensive explanation was showed in the Supplementary Materials for the correction for matching errors caused by body position changes. During the reuse of MARNS in surgery, another sterilized QR code iron plate was positioned in predetermined locations for rescanning, in order to reconfirm the precise location of the tumor for the procedure.
In the x-ray group, a conventional x-ray scan was performed by Dr. Tian Li, a radiologist with 4 years of experience, to determine the level of the target spine. Prior to opening the spinal canal, Dr. Li reverified the spinal level using an x-ray for a second time to confirm the precise location of the surgical site, ensuring the highest level of precision and safety during the procedure.

5. Positioning Time

In the MARNS group, the duration of spinal positioning encompassed the process of placing a QR code iron plate, determining the level of the intended spinal region and its corresponding dorsal skin projection, as well as revalidating the positioning of the target spinal level throughout the surgical procedure.
For patients undergoing x-ray radiation, the duration of spinal positioning entailed the placement of the C-arm machine, determination of the target spinal level and its corresponding back skin projection, and subsequent reconfirmation of the positioning of the target spinal level during the operation.

6. Intraoperative Incision Variation

The double error theory we proposed, indicated that neglecting spinal curvature, which may influence incision placement, constitutes the primary cause of spinal leveling error. Consequently, we employ incision length variation, defined as the difference between the designed incision size and the intraoperative incision size, as an indicator to assess the surgical impact of such error. For the designed incision length design, we centered it on the projection points and then extended it approximately 2 cm upwards and downwards, respectively (Fig. 2D). Slight variations in length were considered normal as the actual incision made during surgery was not always perfectly aligned with the predesigned incision. However, when the target spinal canal level significantly deviated from its central position, making the initial incision inadequate for full lesion exposure, it became necessary to extend the incision. This extension was crucial to adequately expose the spinal canal and clearly visualize the upper and lower boundaries of the intraspinal lesion, thereby ensuring surgical precision and safety.

7. Accuracy Analysis

The evaluation of the positioning outcome involved assessing the agreement between the intraoperative localization of the spinal canal and its actual level. Accuracy was determined by categorizing the alignment with the lesion-related spinous process segments into 3 tiers [11]: (1) Accurate was defined as an alignment with no perceptible deviations. (2) Minor deviation referred to a noticeable divergence, yet not quite reaching one complete segment. (3) Major deviation signifies a substantial misalignment of one or more complete segments from the presumptive level.

8. Error Analysis of Positioning With MARNS

During intraoperative procedures utilizing the MARNS for tumor localization, surgeons often rely on the 3D tumor model generated by the system to determine the tumor’s boundaries. To assess the accuracy of the MARNS system, surgeons measured the distance between the estimated positions provided by the model and the actual tumor location. This measurement serves as a quantifiable indicator of localization error. However, in the case where the tumor is located within the spinal cord and its precise localization cannot be determined prior to surgical exposure, surgeons must employ alternative methods. This study used a surgical instrument to physically indicate the tumor’s boundaries. Following this indication, surgeons used screenshots that included the QR code metal patch of known dimensions (2.3 cm in length), which was placed nearby for reference (Fig. 2E). By comparing the relative positions of the tumor boundaries and the QR code metal patch in the screenshot, we could calculate the distance between the estimated tumor boundaries from the 3D model and the actual tumor location.

9. Data Collection and Definition

Clinical data, including patients’ age, sex, body mass index (BMI), spinal level of the intraspinal lesion, classification, pathologic diagnosis, and lesion size, were obtained from medical records reports. Intraoperative and postoperative information was also collected, including positioning time, the length of designed and intraoperative incision, and accuracy analysis of lesion localization.

10. Statistical Analysis

The statistical analysis was performed using IBM SPSS Statistics ver. 25.0 (IBM Co., Armonk, NY, USA). Continuous data with a normal distribution were described as mean±standard deviations and analyzed using Student t-test. Categorical data were described as n (%) and analyzed using the chi-square or Fisher exact test. Two-sided p-values <0.05 were considered statistically significant.

RESULTS

A total of 41 patients (19 males) with intraspinal tumors were included, and MARNS was applied to 21 patients. There were no significant differences in age, sex, BMI, level of the intraspinal lesion, classification, pathologic diagnosis, or lesion size between MARNS and x-ray groups (all p>0.05) (Table 1, Supplementary Tables 1 and 2).
Accurate localization of intraspinal lesions was done for all patients in the MARNS (Fig. 2F) and conventional x-ray group. However, the positioning time was significantly shorter in the patients receiving MARNS than in the x-ray group (129.00±13.03 seconds vs. 365.00±60.43 seconds, p<0.001). The MARNS system demonstrated a localization error of 0.38±0.12 cm (Fig. 5).
In comparing the incision length variation, the MARNS group had a designed average of 7.33±1.62 cm and an intraoperative average of 7.48±1.55 cm, resulting in an increase of 0.14 cm. Conversely, the designed incision of the x-rays group averaged 6.85±1.15 cm, but the intraoperative length increased to 7.52±1.68 cm, an increase of 0.67 cm. The incision variation between groups was statistically significant (p=0.009) (Table 2). The designed and intraoperative incision length of every patient are shown in Fig. 6. While the localization of the target spinal level in the conventional x-ray group proved accurate, secondary verification revealed that, in 4 cases, the target level deviated from the intended incision center, necessitating larger incisions.

DISCUSSION

This study highlights how surgeons’ neglecting spinal curvature and visual habits can lead to double positioning errors in spinal surgeries. To tackle this, we introduced the self-developed MARNS system. Using AR technology, it vertically projects spinal canal lesions onto the patient’s back, reducing the impact of the double errors on horizontal spinal localization. MARNS offers precise, fast localization and accurate incisions compared to x-rays. However, the time taken for the positioning step does not account for preoperative steps like 3D model reconstruction and surgical planning which are crucial but somewhat laborious. Ultimately, the study mainly shows MARNS’s advantage of eliminating ionizing radiation exposure. In brief, MARNS is a potentially effective tool for locating lesions within the spinal canal.
What are the reasons for spinal surgery performed at the wrong level? Watts et al. [20] analyzed Veterans Health Administration data from 2000–2017, finding 32 wrong-site spine surgeries often due to radiograph issues. A survey [21] on spine localization techniques reported that 77% of surgeons who made errors commented on their mistakes, which fell into preoperative and intraoperative categories. Preoperative errors were often due to imaging issues, while intraoperative errors stemmed from communication gaps, failure to reposition, insufficient reference points, and counting errors.
Although intraspinal tumors can be accurately located and verified by various methods, significantly reducing the incidence of wrong-level spinal surgeries, discrepancies may still occur. [22]. Naqvi et al. [23] assessed the current practices of spinal surgeons across the United Kingdom concerning the techniques implemented for correct-level verification and found that 47.5% of the 105 spine surgeons were involved in wrong-level spinal surgery. Many surgeons reported situations where the spinal level identified pre-incision differed from the postincision exposure, a significant concern during surgical procedures. This underestimated yet serious issue, known as unintentional vertical exposure, arises in 1.3% to 15% of spinal surgeries [24,25], and often requires an extension of the incision length. This elongation not only extends the duration of the operation and complicates the surgical process but also necessitates more extensive dissection of soft tissues [26]. As a result, patients are exposed to heightened surgical risks and potentially more challenging postoperative recovery trajectories.
Notwithstanding the above-mentioned, the specific cause of incorrect-level spine surgeries remains unknown. In our study, although initial x-ray localization appeared accurate, intraoperative verification showed that in several cases, pre- and postincision spinal level assessments did not align, necessitating larger cuts. This deviation aligns with our proposed dual-error theory, suggesting the radiologist may have overlooked spinal curvatures, compounded by the surgeon’s unconscious visual biases, thereby amplifying localization errors of incision and even causing unintentional vertical exposure. Surprisingly, few studies have documented this crucial factor.
AR technology is a promising solution to address the challenge of positional deviation stemming from the inherent physiological curvature of the spine and the surgeon’s visual habits. Although AR navigation research is still in its early stages [27], it offers superior precision and efficacy in pedicle nail implantation and percutaneous vertebroplasty compared to fluoroscopyassisted navigation [28]. By integrating real-time, 3D visualization and precise spatial orientation, AR technology can enhance surgical accuracy, reduce complications, and optimize outcomes [29]. In this study, with the help of visualization capabilities of AR technology, we orthogonally projected a 3D lesion model onto the patient’s actual back surface. This allowed for the accurate ascertainment of the precise spinal level for all patients, resulting in no unintentional vertical exposure or wrong-level surgery in the MARNS group. In the preparation of MARNS, although patients are positioned prone during CT scans, similar to the surgical position, intraspinal tumors may still undergo displacement due to slight variations in body positions during surgery. However, because of the relatively rigid structural characteristics of the spine, the variation in their relative location along the longitudinal axis is minimal and can be largely ignored. Therefore, according to our proposed dual-error theory, to determine the incision center, it suffices to vertically project the 3D tumor model generated by MARNS onto the patient’s actual dorsal body surface, rather than onto a virtual surface, ensuring accurate incision placement (see Supplementary Materials for details).
X-ray radiation poses a potential risk to patients and medical professionals in the operating room. The assessment of setup radiation in the context of total radiation exposure during spine minimally invasive surgery revealed a significant rise in radiation exposure during preoperative localization [30,31]. This finding underscores the persistent risks associated with radiation exposure during surgical procedures despite the implementation of multiple safeguards. It is especially pertinent given the increased susceptibility of medical personnel, who often face frequent exposure and potentially experience long-term health consequences. The pressing need for enhanced protective measures to ensure the well-being of patients and medical staff has led us to explore the application of MARNS.
Recent studies have demonstrated that AR technology can facilitate spine surgery with reduced radiation exposure to patients and the treatment team. Elmi-Terander et al. [32] conducted a study that involved 20 patients undergoing AR-assisted pedicle screw placement in the thoracic and lumbosacral spine. Their results demonstrated acceptable accuracy without the use of fluoroscopic guidance. Meanwhile, Carl et al. [33] found that lowdose intraoperative CT can resolve AR registration errors by <1 cm and reduce radiation exposure by 70%, thereby facilitating the use of AR in spine surgery. In this study, the self-developed MARNS was used to facilitate visualization of intraspinal tumor positioning and surgical incision placement, achieving precise outcomes and eliminating the need for x-ray fluoroscopy. This groundbreaking methodology offers a completely radiation-free approach, effectively eliminating the risks linked to radiation exposure.
Additionally, this MARNS system distinguishes itself by its cost-effectiveness, a particularly noteworthy attribute given the prevailing high costs [28] associated with AR spine positioning systems and the resultant barriers to their widespread adoption. The MARNS capitalizes on mobile AR technology via a smartphone application platform, obviating the necessity for specialized and expensive equipment. This approach not only enhances the accessibility of our system but also substantially alleviates the financial burden, thereby ensuring its affordability for a wider array of healthcare facilities and practitioners.
In summary, this study explores a radiation-free, and cost-effective localization method using AR technology integrated with smartphones for precise intraspinal tumor surgeries. Additionally, we introduce the dual-error theory, presenting surgical errors from neglecting spinal curvature and visual habits, and leverage AR’s visualization for effective spinal level localization. The method also serves as a precise complement to traditional localization techniques, especially when C-arm and similar devices are limited or unavailable, affording spinal surgery accuracy.
However, this study has limitations. Firstly, as a retrospective study, relying on incision length variation for evaluation has its limitations due to potential surgeon bias, which may skew the results. Additionally, a larger sample size would be beneficial. Clearly, there is a need for carefully designed large-scale multicenter studies. Secondly, while MARNS eliminates intraoperative radiation, preoperative CT exams are still required, potentially increasing radiation exposure and cost for patients. Thirdly, MARNS is tailored for intraspinal tumor surgeries and may not suit pedicle screw or other spinal surgeries without adaptation. Lastly, to achieve the goal of creating a radiation-free localization system and overcome the matching discrepancies between preoperative images and the real world, thereby improving the positioning accuracy and expanding the application scope of MARNS, it is necessary to explore image deformation techniques. This will enable us to reach a seamless alignment between the virtual and real worlds. Additionally, we recognize the potential contamination risks posed by using mobile phones as display devices during surgery. Future research will focus on exploring advanced equipment like head-mounted displays or AR glasses to enhance surgical convenience and safety. These devices show promise in maintaining surgical sterility and deserve further investigation prior to clinical use.

CONCLUSION

By utilizing AR technology to visualize the lesion and its perpendicular projection onto the dorsal body surface, MARNS facilitates precise localization of the lesion and the planned incision. Additionally, MARNS operates without intraoperative radiation and requires only basic consumer-grade materials, specifically a mobile smartphone and 2 QR code metal sheets. However, multicenter prospective studies with larger sample sizes are still crucial to validate and enhance the applicability of this innovative approach.

Supplementary Material

Supplementary Material, Supplementary Tables 1-2, Supplementary Fig. 1, and Supplementary video clips 1-2 can be found via https://doi.org/10.14245/ns.2448188.094.
Supplementary Materials.
ns-2448188-094-Supplementary-Material.pdf
Supplementary Table 1.
Record of lesion characteristics and surgical incision data for the group MARNS
ns-2448188-094-Supplementary-Table-1.pdf
Supplementary Table 2.
Record of lesion characteristics and surgical incision data for the group x-rays
ns-2448188-094-Supplementary-Table-2.pdf
Supplementary Table 3.
Calculation of the actual displacement distance of the MARNS 3D model of lesions
ns-2448188-094-Supplementary-Table-3.pdf
Supplementary Fig. 1.
Derived diagram of actual displacement distance of MARNS 3D tumor model. (A) The MARNS displayed a 3D model that exhibited slight deviations from the actual scenario during the matching process. (B) The solid contour a represents the body surface and spine of the patient in the real world; the dashed line b represents the body surface and spine of the virtual 3D model; the angle ∠β between the 3D body surface model and the actual situation can be measured with the protractor measurement program (China, V2.12). Where point O is the intersection between the 3D model and the actual body surface contour, point O’ is the intersection point between the 3D spine model and actual the spine contour, point T’ is the 3D tumor model, point T is perpendicular projection on the actual body surface of point T’, point T’’ is to the body surface projection of point T’ displayed in vertical view with MARNS, point t’ represents where the tumor is, and point t represents a perpendicular projection on the actual body surface of point t’. (C) A geometric representation of diagrams A and B. MARNS, mobile augmented reality navigation system; 3D, 3-dimentional.
ns-2448188-094-Supplementary-Fig-1.pdf
Supplementary video clip 1.
Supplementary video clip 2.

NOTES

Conflict of Interest

The authors have nothing to disclose.

Funding/Support

This study was supported by the Fujian Provincial Science and Technology Department guided project (2023Y0048), the Fujian Provincial Science and Technology Department Foreign Cooperation Project (2020I0028), and the Fujian Medical University Sailing Fund Project (No. 2019QH1111).

Author Contribution

Conceptualization: TL, YL; Data curation: WH, TL, JL, YL, ZC, BH, YW, YL; Formal analysis: WH, JL, YL, SH, BH, YW, YL; Funding acquisition: WH; Methodology: XH, SH, ZC; Writing – original draft: WH; Writing – review & editing: XH, BH, YW, YL.

Fig. 1.
Schematic representation of the supposition regarding incorrect-level spine surgery. (A) The tumor is shown in solid green; the projection of the tumor on the body surface is shown in yellow, the incision in red, the surgical field range in blue, and the surgeon’s estimation of the tumor location is depicted in dotted green. (B) A geometric model from the left graph collated using D=h × sinα, where ∠α is the oblique angle of the incision, h represents the distance from the back surface to the tumor, and D signifies the discrepancy between the actual and estimated tumor locations.
ns-2448188-094f1.jpg
Fig. 2.
The application process of MARNS. (A) The upper and lower poles of the tumor (in blue) were vertically projected (red arrows) onto the dorsal body surface while reconstructing the 3-dimentional (3D) virtual model. (B, C) The 3D model exhibited by MARNS showing (B) a vertical view and (C) a lateral view. (D) Tumor’s projection and incision were determined using a 3D model. (E) The projection of the 3D lesion model intraoperatively. (F) The tumor was exposed at the predetermined location. MARNS, mobile augmented reality navigation system.
ns-2448188-094f2.jpg
Fig. 3.
Registration of spinal computed tomography and magnetic resonance imaging.
ns-2448188-094f3.jpg
Fig. 4.
The development process of MARNS application. CT, computed tomography; MRI, magnetic resonance imaging; STL, stereolithography; FBX, filmbox; AR, augmented reality; APK, android package kit.
ns-2448188-094f4.jpg
Fig. 5.
Error analysis of positioning with mobile augmented reality navigation system (MARNS).
ns-2448188-094f5.jpg
Fig. 6.
Tumor sizes and comparison of designed and intraoperative incision lengths. MARNS, mobile augmented reality navigation system.
ns-2448188-094f6.jpg
Table 1.
Patient’s characteristics
Characteristic MARNS (n = 21) Conventional X-rays (n = 20) p-value
Sex 0.758
 Male 9 10
 Female 12 10
Age (yr) 0.683
 < 18 2 1
 18–40 6 8
 41–65 13 11
BMI (kg/m2) 0.837
 18.5–23.9 8 7
 24.0–27.9 13 13
Spinal level of the lesions* 0.461
 Cervical (C) 1 2
 Cervicothoracic 0 1
 Thoracic (T) 9 10
 Thoracolumbar 2 0
 Lumbar (L) 9 7
Classification of lesions* 0.326
 Extradural 2 0
 Intradural-extramedullary 15 17
 Intramedullary lesions 4 3
Pathologic diagnosis 0.824
 Neurilemmoma 13 13
 Meningioma 3 2
 Neurofibroma 1 2
 Glioma 2 1
 Ependymoma 1 2
 Myeloma 1 0
Tumor size (cm) 3.19 ± 1.72 2.75 ± 1.14 0.348

Values are presented as number or mean±standard deviation.

MARNS, mobile augmented reality navigation system; BMI, body mass index.

* Intraspinal space-occupying lesions.

Table 2.
Position outcomes between MARNS and x-rays groups
Variable MARNS (n = 21) Conventional x-rays (n = 20) p-value
Positioning time (sec) 129 ± 13.03 365 ± 60.43 < 0.001
Incision length variation (cm) 0.14 ± 0.22 0.67 ± 0.86 0.009
The distance between the localized tumor boundary and the actual tumor location (cm) 0.38 ± 0.12
Accuracy
 Accurate 21 20
 Minor deviation 0 0
 Major deviation 0 0

Values are presented as mean±standard deviation or number.

MARNS, mobile augmented reality navigation system.

REFERENCES

1. Weber C, Gulati S, Jakola AS, et al. Incidence rates and surgery of primary intraspinal tumors in the era of modern neuroimaging: a national population-based study. Spine (Phila Pa 1976) 2014;9:E967-73.

2. Ledbetter LN, Leever JD. Imaging of intraspinal tumors. Radiol Clin North Am 2019;57:341-57.
crossref pmid
3. Goodwin ML, Buchowski JM, Schwab JH, et al. Spinal tumors: diagnosis and treatment. J Ame Acad Orthop Surg 2022;30:e1106-21.
crossref
4. Prod’homme M, Grasset D, Lecocq M, et al. Intraoperative disc level marking with needle: a technical note and prospective study on 30 patients. J Spine Surg 2021;7:190-6.
crossref pmid pmc
5. Longo UG, Loppini M, Romeo G, et al. Errors of level in spinal surgery: an evidence-based systematic review. J Bone Joint Surg Br 2012;94:1546-50.
pmid
6. Grimm B, Laxer E, Blessinger B, et al. Wrong-level spine surgery. JBJS Rev 2014;2:e2.
crossref
7. Bagian JP, Lee C, Gosbee J, et al. Developing and deploying a patient safety program in a large health care delivery system: you can’t fix what you don’t know about. Jt Comm J Qual Improv 2001;27:522-32.
crossref pmid
8. Strong MJ, Santarosa J, Sullivan TP, et al. Pre- and intraoperative thoracic spine localization techniques: a systematic review. J Neurosurg Spine 2022;36:792-9.
crossref pmid pmc
9. Hsiang J. Wrong-level surgery: a unique problem in spine surgery. Surg Neurol Int 2011;2:47.
pmid pmc
10. Francis T, Benzel E. Wrong level spine surgery: a perspective. World Neurosurg 2013;79:451-2.
crossref pmid
11. Jhawar BS, Mitsis D, Duggal N. Wrong-sided and wrong-level neurosurgery: a national survey. J Neurosurg Spine 2007;7:467-72.
crossref pmid
12. Virk S, Qureshi S. Navigation in minimally invasive spine surgery. J Spine Surg 2019;5(Suppl 1):S25-30.
crossref pmid pmc
13. Carl B, Bopp M, Saß B, et al. Augmented reality in intradural spinal tumor surgery. Acta Neurochir (Wien) 2019;61:2181-93.
crossref pdf
14. Carl B, Bopp M, Chehab S, et al. Preoperative 3-dimensional angiography data and intraoperative real-time vascular data integrated in microscope-based navigation by automatic patient registration applying intraoperative computed tomography. World Neurosurg 2018;113:e414-25.
crossref pmid
15. Liu Y, Lee MG, Kim JS. Spine surgery assisted by augmented reality: where have we been? Yonsei Med J 2022;63:305-16.
crossref pmid pmc pdf
16. Edström E, Burström G, Nachabe R, et al. A novel augmented-reality-based surgical navigation system for spine surgery in a hybrid operating room: design, workflow, and clinical applications. Oper Neurosurg (Hagerstown) 2020;18:496-502.
crossref pmid pdf
17. Leger E, Reyes J, Drouin S, et al. MARIN: an open-source mobile augmented reality interactive neuronavigation system. Int J Comput Assist Radiol Surg 2020;15:1013-21.
crossref pmid pdf
18. Shah M, Halalmeh DR, Sandio A, et al. Anatomical variations that can lead to spine surgery at the wrong level: part I, cervical spine. Cureus 2020;12:e8667.
crossref pmid pmc
19. Shah M, Halalmeh DR, Sandio A, et al. Anatomical variations that can lead to spine surgery at the wrong level: part II thoracic spine. Cureus 2020;12:e8684.
crossref pmid pmc
20. Watts BV, Rachlin JR, Gunnar W, et al. Wrong site spine surgery in the veterans administration. Clin Spine Surg 2019;32:454-7.
crossref pmid
21. Mayer JE, Dang RP, Duarte Prieto GF, et al. Analysis of the techniques for thoracic- and lumbar-level localization during posterior spine surgery and the occurrence of wrong-level surgery: results from a national survey. Spine J 2014;14:741-8.
crossref pmid
22. Kumar N, Tan WLB, Wei W, et al. An overview of the tumors affecting the spine-inside to out. Neurooncol Pract 2020;7(Suppl 1):i10-7.
crossref pmid pmc pdf
23. Naqvi AZ, Magill H, Anjarwalla N. Intraoperative practices to prevent wrong-level spine surgery: a survey among 105 spine surgeons in the United Kingdom. Patient Saf Surg 2022;16:6.
crossref pmid pmc pdf
24. Wiese M, Krämer J, Bernsmann K, et al. The related outcome and complication rate in primary lumbar microscopic disc surgery depending on the surgeon’s experience: comparative studies. Spine J 2004;4:550-6.
crossref pmid
25. Ammerman JM, Ammerman MD, Dambrosia J, et al. A prospective evaluation of the role for intraoperative x-ray in lumbar discectomy. Predictors of incorrect level exposure. Surg Neurol 2006;66:470-3. discussion 473-4.
crossref pmid
26. Patel A, Runner RP, Bellamy JT, et al. A reproducible and reliable localization technique for lumbar spine surgery that minimizes unintended-level exposure and wrong-level surgery. Spine J 2019;19:773-80.
crossref pmid
27. Burström G, Persson O, Edström E, et al. Augmented reality navigation in spine surgery: a systematic review. Acta Neurochir (Wien) 2021;163:843-52.
crossref pmid pmc pdf
28. Yuk FJ, Maragkos GA, Sato K, et al. Current innovation in virtual and augmented reality in spine surgery. Ann Transl Med 2021;9:94.
crossref pmid pmc
29. Ghaednia H, Fourman MS, Lans A, et al. Augmented and virtual reality in spine surgery, current applications and future potentials. Spine J 2021;21:1617-25.
crossref pmid
30. Reiser EW, Desai R, Byrd SA, et al. C-arm positioning is a significant source of radiation in spine surgery. Spine (Phila Pa 1976) 2017;42:707-10.
crossref pmid
31. Nelson EM, Monazzam SM, Kim KD, et al. Intraoperative fluoroscopy, portable X-ray, and CT: patient and operating room personnel radiation exposure in spinal surgery. Spine J 2014;4:2985-91.
crossref
32. Elmi-Terander A, Burström G, Nachabé R, et al. Augmented reality navigation with intraoperative 3D imaging vs fluoroscopy-assisted free-hand surgery for spine fixation surgery: a matched-control study comparing accuracy. Sci Reps 2020;10:707.
crossref pmid pmc pdf
33. Carl B, Bopp M, Saß B, et al. Reliable navigation registration in cranial and spine surgery based on intraoperative computed tomography. Neurosurg Focus 2019;47:E11.
crossref


Editorial Office
Department of Neurosurgery, CHA Bundang Medical Center,
CHA University School of Medicine,
59 Yatap-ro, Bundang-gu, Seongnam 13496, Korea
Tel: +82-31-780-1924  Fax: +82-31-780-5269  E-mail: support@e-neurospine.org
The Korean Spinal Neurosurgery Society
#407, Dong-A Villate 2 Town, 350 Seocho-daero, Seocho-gu, Seoul 06631, Korea
Tel: +82-2-585-5455  Fax: +82-2-2-523-6812  E-mail: ksns1987@gmail.com
Business License No.: 209-82-62443

Copyright © The Korean Spinal Neurosurgery Society.

Developed in M2PI

Zoom in Close layer