Clinical application research of navigation system-assisted vertebroplasty for the treatment of lumbar osteoporotic compression fractures

Abstract

Objective

This study aims to compare the clinical efficacy and radiological outcomes of the computer navigation-assisted system versus traditional fluoroscopy-assisted percutaneous vertebroplasty (PVP) in the treatment of lumbar osteoporotic vertebral compression fractures (OVCFs). The goal is to clarify the advantages and disadvantages of the computer navigation system in assisting PVP surgery, thereby providing references and experience for the promotion and application of this technology in minimally invasive spinal surgeries.

Methods

This study is a single-center, prospective, randomized controlled trial. Patients who underwent lumbar percutaneous vertebroplasty (PVP) for the treatment of lumbar osteoporotic vertebral compression fracture (OVCF) at General Hospital of Northern Theater Command from January 2023 to January 2024 were included. Cases that met the inclusion and exclusion criteria were randomly divided into two groups using a random number table method: the computer navigation-assisted PVP group (navigation group) and the traditional fluoroscopy-assisted PVP group (traditional group), with allocation concealment implemented. The following data were recorded for both groups: basic baseline characteristics of the patients; the number and duration of needle punctures, total number and duration of fluoroscopy, and total surgical time; the amount of bone cement used, and the incidence of complications (bone cement leakage rate, nerve injury, vascular injury, and vascular embolism); preoperative and postoperative Visual Analogue Scale (VAS) scores and Oswestry Disability Index (ODI) scores, as well as the mean height of fractured vertebrae (HFV). Follow-up was conducted at 3, 6, and 12 months postoperatively for both groups of patients.

Results

This study recruited 68 patients with lumbar osteoporotic vertebral compression fracture (OVCF) who underwent percutaneous vertebroplasty (PVP) from January 2023 to January 2024. Cases that met the inclusion and exclusion criteria were randomly divided into two groups using a random number table method: the computer navigation-assisted PVP group (navigation group, 35 cases) and the traditional fluoroscopy-assisted PVP group (traditional group, 33 cases), with allocation concealment implemented. In the navigation group, 3 patients were lost to follow-up (loss rate of 8.57%), and in the traditional group, 3 patients were lost to follow-up (loss rate of 9.09%). Finally, the navigation group included 32 cases, and the traditional group included 30 cases. There were no significant differences between the two groups in terms of age, preoperative T-score, and preoperative Visual Analogue Scale (VAS) scores (P > 0.05), indicating comparability. The VAS scores of patients in the navigation group were lower than those in the traditional group when puncturing to the target site. The total number of fluoroscopies in the navigation group was lower than that in the traditional group. The total fluoroscopy time in the navigation group was less than that in the traditional group. The amount of bone cement used in the navigation group was higher than that in the traditional group. The number of bone cement leakage cases in the navigation group was lower than that in the traditional group (P < 0.05). The number of punctures in the navigation group was significantly lower than that in the traditional group, and the puncture time in the navigation group was significantly shorter than that in the traditional group (P < 0.001). No significant vascular injury, nerve injury, or vascular embolism occurred in either group. There were no significant differences in postoperative VAS and ODI scores or in the recovery of the mean height of fractured vertebrae (HFV) between the two groups (P > 0.05).

Conclusion

The computer-assisted navigation systems in PVP can improve the accuracy of needle puncture and surgical safety, reduce the number of fluoroscopic exposures, shorten fluoroscopy time, and decrease radiation exposure for both patients and medical staff during the procedure. This enhances the procedural experience for patients and achieves “visualization of surgical operations”, thereby realizing the goal of “precision medicine”.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12893-026-03505-y.

Introduction

Osteoporosis is a common systemic skeletal disorder characterised by reduced bone mass, impaired microstructural integrity, diminished bone strength, and increased bone fragility, leading to heightened risk of fractures throughout the body [1–3]. As China enters an ageing society, the prevalence of osteoporosis among the elderly continues to rise. Survey data indicates that in 2010, approximately 2.33 million osteoporotic fractures occurred annually, with women experiencing fractures at roughly three times the rate of men. Projections suggest that by 2035, the annual incidence of osteoporosis-related fractures will double, reaching 5.99 million cases by 2050 [4]. For elderly individuals with osteoporosis, osteoporotic vertebral compression fractures (OVCFs) may occur in daily life regardless of a history of trauma, with the likelihood increasing with age. OVCFs frequently cause persistent low back pain with localised radiating pain, restricted mobility, progressive kyphotic deformity, and delayed or non-union of fractures, thereby diminishing patients’ quality of life. They increasingly represent a public health concern threatening elderly health [5]. Current treatment approaches for lumbar OVCFs primarily comprise conservative management and surgical intervention. Conservative treatment involves bed rest, oral analgesics, topical ointments, and orthotic support [6]. Strict long-term bed rest is challenging to implement and prone to complications (e.g., pressure ulcers, pulmonary infections, atelectasis, urinary tract infections, deep vein thrombosis in the lower limbs), potentially leading to non-union or worsening of fractures. Incomplete statistics indicate that mortality rates among long-term bedridden patients can reach 20%, with lifelong disability rates as high as 50% [7]. In 1987, Galibert first introduced vertebroplasty [8], involving the injection of medical bone cement (poly(methyl methacrylate)) into the vertebral body to treat cervical vertebral haemangiomas. Percutaneous Vertebroplasty (PVP) and Percutaneous Kyphoplasty (PKP) have progressively been applied in clinical management of OVCFs with favourable outcomes, now serving as primary minimally invasive techniques for treating these conditions [9, 10]. These techniques offer advantages including minimal trauma, rapid pain relief, restoration of vertebral height and spinal stability, improvement or prevention of kyphotic deformity, increased bone strength, early mobilisation of patients, and prevention of complications associated with prolonged bed rest [11, 12]. Traditional PVP techniques rely on repeated intraoperative C-arm fluoroscopy to confirm needle placement and establish optimal bone cement infusion pathways, yet present several issues: (1)Achieving ideal needle positioning and bone cement pathways necessitates repeated anteroposterior and lateral fluoroscopy, significantly prolonging puncture time and total surgical time; (2)Repeated anteroposterior and lateral fluoroscopy not only prolongs puncture time but also increases intraoperative radiation exposure and dose for both patients and medical staff. Studies indicate that prolonged, frequent low-dose radiation exposure may elevate risks of radiation-induced cataracts and thyroid cancer [13, 14]; (3)༉Suboptimal puncture pathways may lead to spinal nerve injury and secondary bone cement leakage. Among these, bone cement leakage is the most common complication of PVP surgery, occurring in approximately 11%–76% of cases [15]. While some instances of leakage present without obvious clinical symptoms, it may lead to severe consequences including spinal cord and nerve root compression, neurological dysfunction, paraplegia, pulmonary embolism, and vascular embolism [16, 17]. The precision and safety of needle placement depend not only on repeated fluoroscopic guidance to identify optimal puncture sites and needle trajectories but also correlate positively with the surgeon’s experience. In summary, there is an urgent need for a more precise and safer surgical assistance technique to reduce radiation exposure, enhance procedural accuracy, and lower complication rates. With the rise of the ‘precision medicine’ concept, spinal navigation technology—a key application in orthopaedic precision medicine—has matured alongside advancements in stereotactic techniques, image registration, robotics, and computer technology [18]. Current studies indicate [19–21] that computer navigation-assisted PVP offers advantages, potentially doubling efficiency. However, few investigations have compared the clinical application and outcomes of computer navigation-assisted PVP versus conventional fluoroscopy-guided PVP for treating lumbar osteoporotic vertebral compression fractures (OVCFs). In this study, we employed a prospective randomised controlled trial to compare the clinical efficacy and radiographic outcomes of the computer navigation-assisted percutaneous vertebroplasty versus traditional fluoroscopy-guided PVP for treating lumbar osteoporotic compression fractures. The objective is to clarify the advantages and disadvantages of applying this computer navigation system to assist PVP, thereby providing reference and experience for promoting its application in minimally invasive spinal surgery.

Methods

Inclusion and exclusion criteria

Inclusion criteria for subjects

(1) Age ≥ 60 years; (2) T-score≤−2.5 SD; (3) Single-segment lumbar osteoporotic vertebral compression fracture (OVCF); (4) Preoperative three-dimensional lumbar CT demonstrating intact posterior vertebral wall; (5) Preoperative lumbar MRI revealing fresh OVCF or non-union of old fracture.

Exclusion criteria for subjects

(1) Vertebral compression exceeding 75% of vertebral height; (2) Degenerative scoliosis with rotational deformity of the fractured vertebra (Cobb angle > 25°); (3) Pathological vertebral fractures (associated with spinal tumours, infections, tuberculosis, brucellosis, etc.); (4) Presence of lower limb neurological symptoms; (5) Patients with severe obesity (BMI ≥ 30 kg/m²); (6) Systemic diseases or inability to tolerate surgery:①Severe haemorrhagic disorders, coagulation dysfunction, lower limb deep vein thrombosis;②Severe cardiac conditions;③Severe respiratory diseases;④Inability to tolerate anaesthesia or cooperate during surgery.

General information

This study was a single-centre, prospective, randomised controlled trial. Patients undergoing percutaneous vertebroplasty (PVP) for lumbar osteochondral fracture (OVCF) at General Hospital of Northern Theater Command between January 2023 and January 2024 were enrolled. Cases meeting inclusion and exclusion criteria were randomly assigned using a random number table to either the computer navigation-assisted PVP group (navigation group) or the conventional fluoroscopy-guided PVP group (traditional group), with allocation concealed. There were no statistically significant differences between the two groups in terms of gender, age, height, weight, preoperative T-score, preoperative VAS score, preoperative ODI score, aetiology (presence or absence of a clear history of trauma), underlying conditions (hypertension, diabetes mellitus, coronary heart disease), or smoking status (P > 0.05), rendering them comparable. Baseline characteristics are presented in Table 1. Prior to surgery, both patients and their families received comprehensive explanations regarding the surgical approach, perioperative risks, and potential complications. All parties indicated full understanding and informed consent was obtained through signed surgical consent forms. This study was approved by the Ethics Committee of General Hospital of Northern Theater Command. Ethics approval number: Lun Shen Y (2021) No. 025. All procedures were performed by the same surgeon, who had undergone systematic training in computer navigation equipment operation and possessed experience in spinal surgery. Preoperative investigations included lumbar X-rays (anterior-posterior and lateral views), three-dimensional lumbar CT, lumbar MRI, electrocardiogram, cardiac echocardiography, and bilateral lower limb arterial and venous Doppler ultrasound. Postoperative imaging assessments were conducted independently by two physicians with extensive spinal surgery experience but no prior exposure to computer navigation technology, alongside one senior radiologist, using a double-blind approach. All discrepancies were resolved through discussion and consensus.

Table 1.

Comparison of baseline characteristics between the two groups of patients

Group	Navigation group  (n=32)	Traditional group  (n=30)	χ2/t value	P value
Sex	Male（8/32=25.00%） Female（24/32=75.00%）	Male（8/30=26.67%） Female（22/30=73.33%）	0.022	0.881
Age	71.91±7.47	71.77±8.78	0.068	0.946
Height（cm）	162.38±8.28	162.70±7.62	−0.160	0.873
Weight（kg）	63.50±10.54	63.97±11.40	−0.167	0.868
T-value（SD）	−3.16±0.26	−3.13±0.27	−0.434	0.666
VAS score	7.00±1.63	6.90±1.63	0.242	0.810
ODI score	74.47±11.74	72.23±14.05	0.681	0.498
History of trauma	Yes（25/32=78.13%）	Yes（22/30=73.33%）	0.194	0.660
Hypertension	No（18/32=56.25%） Yes（14/32=43.75%）	No（16/30=53.33%） Yes（14/30=46.67%）	0.053	0.818
Diabetes	No（24/32=75.00%） Yes（8/32=25.00%）	No（24/30=80.00%） Yes（6/30=20.00%）	0.221	0.638
Coronary heart disease	No（26/32=81.25%） Yes（6/32=18.75%）	No（25/30=83.33%） Yes（5/30=16.67%）	0.046	0.830
Smoking history	No（26/32=81.25%） Yes（6/32=18.75%）	No（22/30=73.33%） Yes（8/30=26.67%）	0.555	0.456
Type of fracture	Fresh（26/32=81.25%）	Fresh（25/30=83.30%）	0.046	0.830
	Old（6/32=18.75%）	Old（5/30=16.70%）
Distribution of fracture segments n（%）
L1	13（40.625%）	12（40.000%）
L2	11（34.375%）	10（33.333%）
L3	4（12.500%）	3（10.000%）
L4	2（6.250%）	3（10.000%）
L5	2（6.250%）	2（6.667%）

Surgical procedure

The computer navigation cart primarily comprises an infrared camera (NDI) and a central control panel (Fig. 1). The infrared camera (NDI) captures and tracks the real-time position of the optical tracking ball, thereby constructing a three-dimensional spatial model. This ultimately enables real-time feedback of the puncture needle’s position during the procedure. The central control panel primarily facilitates the matching of intraoperative fluoroscopic images with preoperative 3D CT scans of the fractured vertebral body. Upon successful alignment, the displayed real-time imagery enables adjustment of the puncture needle’s position and execution of the puncture procedure along the predefined pathway.

Fig. 1.

Computer navigation cart

Surgical procedure for the navigation group

Phase 1:Preoperative preparation

Preoperatively, a comprehensive 3D CT scan of the patient’s lumbar spine was performed. Imaging data were imported into preoperative planning software, where the lumbar fracture segment was selected and segmented independently. A satisfactory surgical pathway was designed using the three-point alignment method, completing preoperative planning of the puncture trajectory. This was exported for intraoperative three-dimensional data navigation (Fig. 2). Ensure the latest stable installation package is prepared in advance and tested for installation within the central control panel of the mobile unit. Intraoperative C-arms commonly used during procedures are primarily categorised into two types: image-enhanced C-arms and flat-panel C-arms. The methods for correcting these two types of C-arms during the preoperative preparation phase differ. Position the distortion correction plate within the fluoroscopic imaging area of image-enhanced C-arms. Import the fluoroscopic image into the navigation software on the central control panel to perform preoperative distortion correction. Flat-panel C-arms require no such correction and can be used directly for intraoperative fluoroscopic imaging.

Fig. 2.

Preoperative segmentation of fractured vertebral bodies, planning of puncture pathways, and preview

Phase 2:Intraoperative preparation and implementation

The patient is positioned prone with the abdomen suspended. The injured vertebrae are pre-marked, followed by routine disinfection, draping with sterile surgical drapes, and application of skin protection film to the surgical site. Pre-cut skin adhesive film is used to securely affix the patient reference frame to the spinous process region of the first or second vertebrae below the fractured vertebral body (Fig. 3). Four optical tracking spheres are securely fixed to the patient reference frame beforehand. Any loosening may prevent the infrared camera from detecting their positions. Position the trolley at the patient’s caudal end, ensuring it can identify and track the optical tracking spheres. Upon successful recognition of the spheres on the reference frame, the corresponding icon on the central control panel will turn green.

Fig. 3.

Patient reference frame secured with skin adhesive film

Position the image registrar with five optical tracking spheres pre-installed directly above the fractured vertebral body. The image registrar must not obstruct the patient reference frame. The C-arm placement must not conflict with the trolley’s irradiation range. The NDI must be capable of tracking the optical tracking spheres on the navigation instrument during both anteroposterior and lateral fluoroscopy. The first anteroposterior fluoroscopy is performed at the patient’s end-expiratory phase. The navigation system calculates the spatial relationship between the patient’s anatomical position and the surgical instrument using two-dimensional images and optical information. The patient reference frame serves as the skeletal coordinate system; all instruments are spatially calculated relative to this frame using optical and image data. Import the anteroposterior fluoroscopy image into the central console. The navigation software will identify the five optical tracking spheres and six optical markers on the image register. Should positioning be suboptimal, manually adjust the identification location and marker size. Subsequently, position the image register adjacent to the fractured vertebral body in the lateral view, ensuring all five optical tracking spheres and six optical markers are clearly visible within the fluoroscopic image. Employ the same methodology for importing and adjusting lateral fluoroscopic images (Fig. 4). Successful and highly accurate identification of anteroposterior and lateral fluoroscopic images is indicated by a green colouration in the upper left corner of the image (Fig. 5). At this stage, use a probe to identify anatomical landmarks and verify fluoroscopic image accuracy (Fig. 6). Subsequently, perform image position registration between the pre-operatively planned three-dimensional stereotactic fracture vertebra and the two-dimensional images captured during intraoperative anteroposterior and lateral fluoroscopy (Fig. 7). During fluoroscopic image acquisition, the optical tracking spheres, patient reference frame, and image registrator must remain stationary. Movement will result in loss of real-time positioning, leading to final registration failure.

Fig. 4.

Intraoperative C-arm anteroposterior and lateral fluoroscopy

Fig. 5.

Anterior-posterior and lateral fluoroscopic images imported into the central control console (green indicator lights signify high precision)

Fig. 6.

Probe identification of anatomical location and positioning

Fig. 7.

Two-Dimensional and three-Dimensional Image registration

Upon completion of the above steps, all instruments except the patient reference frame may be removed. Pre-position the eight optical tracking spheres onto the needle holders (four) and calibrators (four). Secure the needle firmly to the needle holder, ensuring both the needle holder and calibrator are within the NDI recognition zone. Upon successful recognition, the needle and calibrator indicators on the central console will turn green. Insert the needle tip into the 3.2 mm slot of the calibrator to calibrate the real-time position of the needle tip. Following successful calibration, proceed to select and mark the skin puncture entry point (Fig. 8).

Fig. 8.

Needle position marking and determining the puncture sitet

10 ml of 0.1% lidocaine and 10 ml of 75 mg ropivacaine solution were mixed with normal saline to a total volume of 40 ml. This solution was used for local infiltration anaesthesia of the puncture pathway. Following completion of anaesthesia, the operator may proceed with the puncture according to the pre-planned trajectory. The operator may select one of the following three imaging modes (3D imaging mode, MPR imaging mode, 2D imaging mode) for the puncture procedure, based on personal preference (Fig. 9). During the puncture, when the navigation system indicates the needle tip position on the anteroposterior image is at the medial border of the pedicle and on the lateral image is at the posterior margin of the vertebral body, perform the first anteroposterior and lateral fluoroscopic verification to confirm accuracy. A second anteroposterior and lateral fluoroscopic check is performed when the navigation system indicates the needle tip has reached the pre-planned target site. The entire puncture procedure is thus ‘visualised’, allowing the operator to adjust the needle position in real-time and under direct visual guidance as required (Fig. 10). It is particularly important to note that any instrument deformed due to external force during the procedure must be replaced promptly to avoid compromising surgical precision.

Fig. 9.

Three imaging modes of the navigation system

Fig. 10.

Navigation imaging and fluoroscopic imaging

Upon reaching the target puncture endpoint, remove the needle core, insert the guidewire, withdraw the puncture needle, and position the working sleeve. Decide whether to use a fine drill based on specific circumstances. Thoroughly and rapidly shake the bone cement powder with the bone cement slurry to prepare the liquid bone cement. Inject the bone cement into the bone cement plunger. Proceed with toothpaste-like cement injection along the cement pathway at a rate of 0.3–0.5 ml per push. Cease injection once satisfactory cement placement is confirmed under anteroposterior and lateral fluoroscopy. After cement solidification, remove the cement plunger and apply sterile dressing. Procedure complete.

Surgical procedure for the traditional group

The patient is positioned prone on a bridge-shaped air cushion, with the abdomen suspended. Under C-arm fluoroscopy, the injured vertebral pedicle is localised and marked. The surgical site skin is routinely disinfected and draped. Mix 10 ml of 0.1% lidocaine and 10 ml of 75 mg ropivacaine solution with normal saline to a total volume of 40 ml. Use this prepared anaesthetic solution for local infiltration anaesthesia of the puncture pathway. Under repeated anteroposterior and lateral fluoroscopy, advance the needle through the injured vertebral pedicle to the anterior third of the fractured vertebral body. Subsequently, withdraw the needle core, insert a guide wire, remove the needle cannula, and advance the working cannula along the guide wire. Once satisfactory positioning is confirmed under fluoroscopy, withdraw the guide wire. Alternatively, a fine drill may be slowly advanced under fluoroscopic guidance. Thoroughly and rapidly shake the bone cement powder with the bone cement mixing solution to prepare the liquid bone cement. Inject the bone cement into the bone cement syringe barrel. Along the bone cement channel, inject the toothpaste-like bone cement at a rate of 0.3–0.5 ml per injection. Cease injection when satisfactory cement placement is confirmed under anteroposterior and lateral fluoroscopy. Allow the cement to solidify, then remove the cement plunger. Dress the puncture site with sterile dressings. Procedure complete.

Postoperative and perioperative management

Following the procedure, the patient returned to the ward for 2 h of bed rest, during which close monitoring was maintained of spontaneous lower limb movement and vital signs. Postoperative lumbar spine X-rays (anterior-posterior and lateral views) and three-dimensional CT scans were performed. Systematic anti-osteoporosis treatment was administered throughout the perioperative period.

Follow-up time

Follow-up assessments were conducted at 3, 6, and 12 months post-surgery for both patient groups.

Observation and evaluation indicators

Clinical efficacy indicators

Record the Visual Analogue Scale (VAS) scores for two patient groups at the following time points: preoperatively, puncturing to the target site, postoperatively, and at each follow-up interval. Additionally, record the Oswestry Disability Index (ODI) scores preoperatively, postoperatively, and at each follow-up interval.

Surgical-related parameters

Record the following for both patient groups: number of punctures, puncture time, total number of fluoroscopies, total fluoroscopy time, and total surgical time; bone cement usage, complication incidence (cement leakage rate, vascular injury, nerve injury, vascular embolism).

Puncture to the target site: C-arm intraoperative anteroposterior fluoroscopy indicates the needle tip is positioned within the mid-range between the medial border of the pedicle and the spinous process; lateral fluoroscopy confirms the needle tip has reached the anterior third of the fractured vertebral body.

Number of punctures: The count from initiating the needle insertion to reaching the target site.

Puncture time: Time elapsed from initiation of puncture to reaching the target site.

Total fluoroscopy time: Cumulative duration of C-arm fluoroscopy use.

Total surgical time: Time from commencing sterile draping to completion of bone cement injection (removal of injection rod) and application of sterile dressing.

Imaging-related parameters

Record the mean height of fractured vertebrae (HFV) for two groups at preoperative, postoperative, and each follow-up time point. The mean height of fractured vertebrae (HFV): The mean value of the sum of the anteroposterior heights of the fractured vertebrae measured on lateral radiographs in the neutral position.

Statistical analysis

Statistical analysis of the data was performed using SPSS 26.0 software. When quantitative data conformed to a normal distribution, they were expressed as mean ± standard deviation (x ± s), with intergroup differences analysed using the independent samples t-test. When quantitative data did not conform to a normal distribution, they were expressed as median and interquartile range (M [Q25, Q75]), with intergroup differences analysed using the Wilcoxon signed-rank test. For categorical data, the chi-square test or Fisher’s exact test was applied according to sample size and theoretical frequency. Differences were considered statistically significant at P < 0.05.

Results

General information

This study enrolled 68 patients undergoing percutaneous vertebroplasty (PVP) for lumbar osteochondral fracture (OVCF) at General Hospital of Northern Theater Command between January 2023 and January 2024. Cases meeting inclusion and exclusion criteria were randomly assigned using a random number table to either the computer navigation-assisted PVP group (navigation group, n = 35) or the conventional fluoroscopy-guided PVP group (traditional group, n = 33), with allocation concealed. Three patients were lost to follow-up in the navigation group (loss rate: 8.57%) and three in the traditional group (loss rate: 9.09%). Ultimately, 32 patients were included in the navigation group and 30 in the traditional group. The navigation group comprised 32 patients: 8 males (25.00%) and 24 females (75.00%); The traditional group comprised 30 patients: 8 males (26.67%) and 22 females (73.33%). The mean age in the navigation group was (71.91 ± 7.47) years, while the traditional group averaged (71.77 ± 8.78) years. Patients in the navigation group weighed (63.50 ± 10.54) kg, while those in the traditional group weighed (63.97 ± 11.40) kg. Preoperative T-scores were (−3.16 ± 0.26) SD for the navigation group and (−3.13 ± 0.27) SD for the traditional group. No statistically significant differences were observed between the two groups in baseline characteristics including gender, age, height, weight, preoperative T-score, preoperative VAS score, preoperative ODI score, aetiology (presence or absence of a clear history of trauma), underlying conditions (hypertension, diabetes mellitus, coronary heart disease), or smoking status (P > 0.05). The two groups were comparable (Table 1.).

Clinical efficacy indicators

The preoperative VAS score for patients in the navigation group was (7.00 ± 1.63 points), while that for the traditional group was (6.90 ± 1.63 points) (P = 0.810). There was no statistically significant difference in preoperative VAS scores between the two groups, indicating comparability. At the time of reaching the target site, patients in the navigation group had a VAS score of (6.03 ± 1.43 points), while those in the traditional group had a VAS score of (6.87 ± 1.57 points) (P = 0.032). Patients in the navigation group exhibited a lower VAS score upon reaching the target site compared to the traditional group, with this difference being statistically significant (P < 0.05). Both groups exhibited significant postoperative VAS reductions compared to preoperative levels, though no statistically significant difference existed between them (P > 0.05). At 3-, 6-, and 12-month follow-ups, the navigation group maintained lower VAS scores than the conventional PVP group, yet these differences lacked statistical significance (P > 0.05) (Table 2).

Table 2.

Comparison of VAS scores at different time points between two groups of patients

Group	Navigation group	Traditional group	t value	P value
Preoperative	7.00 ± 1.63	6.90 ± 1.63	0.242	0.810
Puncture to the target site	6.03 ± 1.43	6.87 ± 1.57	−2.196	0.032
1 st day postoperation	3.09 ± 1.20	3.13 ± 1.22	−0.128	0.898
3rd month postoperation	3.03 ± 1.20	3.10 ± 1.18	−0.226	0.822
6 th month postoperation	2.81 ± 1.40	3.03 ± 1.38	−0.625	0.534
12 th month postoperation	2.47 ± 1.32	2.70 ± 1.44	−0.659	0.512

The preoperative ODI score for the navigation group was (74.47 ± 11.74 points), while that for the traditional group was (72.23 ± 14.05 points) (P = 0.498). There was no statistically significant difference in preoperative ODI scores between the two groups, rendering them comparable. Postoperative ODI scores in both groups showed a significant reduction compared with preoperative levels. At different follow-up time points, there were no statistically significant differences in ODI scores between the two groups (P > 0.05) (Table 3).

Table 3.

Comparison of ODI scores at different time points between two patient groups

Group	Navigation group	Traditional group	t value	P value
Preoperative	74.47 ± 11.74	72.23 ± 14.05	0.681	0.498
1 st day postoperation	28.56 ± 3.10	28.60 ± 4.13	−0.041	0.968
3rd month postoperation	26.63 ± 2.27	26.60 ± 2.33	0.043	0.966
6 th month postoperation	21.19 ± 3.11	21.27 ± 3.15	−0.100	0.921
12 th month postoperation	18.41 ± 4.72	18.77 ± 4.63	−0.303	0.763

Surgical-related parameters

The navigation group required 1.00 (1.00, 2.00) punctures, while the traditional group required 4.50 (4.00, 5.25) punctures. The navigation group exhibited significantly fewer punctures than the traditional group, with a statistically significant difference (P < 0.001). The puncture time in the navigation group was (6.91 ± 2.79 min), while that in the traditional group was (10.07 ± 3.32 min). The navigation group exhibited a significantly shorter puncture time than the traditional group, with a statistically significant difference (P < 0.001). The total number of fluoroscopy views in the navigation group (25.28 ± 6.38) was lower than that in the traditional group (29.50 ± 5.98) (P = 0.009), with a statistically significant difference (P < 0.05). The total fluoroscopy time in the navigation group was 20.59 ± 6.72 min, compared to 23.87 ± 5.86 min in the traditional group (P = 0.046). The navigation group exhibited a statistically significant reduction in total fluoroscopy time (P < 0.05). The total surgical time in the navigation group was 26.78 ± 7.21 min, while that in the traditional group was 28.70 ± 7.77 min (P = 0.317). Although the navigation group exhibited a shorter total surgical time than the traditional group, no significant statistical difference was observed (P > 0.05) (Table 4).

Table 4.

Comparison of intraoperative parameters between the two patient groups

Group	Navigation group	Traditional group	t/Z value	P value
Number of punctures M(P25, P75)	1.00(1.00, 2.00)	4.50(4.00, 5.25)	−5.108	P<0.001
Puncture time(min)	6.91 ± 2.79	10.07 ± 3.32	−4.069	P<0.001
Total number of fluoroscopies	25.28 ± 6.38	29.50 ± 5.98	−2.682	0.009
Total fluoroscopy time(min)	20.59 ± 6.72	23.87 ± 5.86	−2.037	0.046
Total surgical time(min)	26.78 ± 7.21	28.70 ± 7.77	−1.008	0.317

The navigated group utilised (6.19 ± 1.37 ml) of bone cement, whereas the traditional group employed (5.45 ± 1.02 ml) (P = 0.020). The navigated group’s cement usage exceeded that of the traditional group, with this difference being statistically significant (P < 0.05). No major vascular injuries, nerve injuries, or vascular embolisms occurred in either group. Bone cement leakage was observed in 2 cases (6.25%) in the navigation group and 9 cases (30.00%) in the traditional group. The incidence of bone cement leakage was lower in the navigation group than in the traditional group, with a statistically significant difference (P < 0.05) (Table 5).

Table 5.

Comparison of complications such as bone cement usage and leakage rates between the two patient groups

Group	Bone cement usage (ml)	Surgical complications
		Vascular injury	Nerve injury	Vascular embolism	Bone cement leakage rate
Navigation group	6.19 ± 1.37	0	0	0	2/32(6.25%)
Traditional group	5.45 ± 1.02	0	0	0	9/30(30.00%)
t/χ2 value	2.389				5.984
P value	0.020				0.014

Imaging-related parameters

The mean preoperative height of the fractured vertebral body (HFV) in the navigation group was (27.43 ± 4.01 mm), while that in the traditional group was (29.04 ± 4.44 mm) (P = 0.137). There was no statistically significant difference in mean preoperative HFV between the two groups. Postoperative HFV levels increased in both groups compared with preoperative values. However, no significant statistical differences were observed between HFV levels at different time points within either group (P > 0.05). At the final follow-up, HFV improved in both groups compared with preoperative levels, yet no significant intergroup differences were noted (P > 0.05) (Table 6).

Table 6.

Comparison of the mean height of fractured vertebrae(HFV) between the two groups of patients

Group	Navigation group(mm)	Traditional group(mm)	t value	P value
Preoperative	27.43 ± 4.01	29.04 ± 4.44	−1.508	0.137
1 st day postoperation	30.92 ± 4.09	31.91 ± 4.16	−0.947	0.347
3rd month postoperation	30.54 ± 3.94	31.55 ± 4.05	−0.989	0.327
6 th month postoperation	30.00 ± 4.11	30.81 ± 4.84	−0.713	0.478
12 th month postoperation	29.25 ± 4.09	30.08 ± 4.39	−0.763	0.449

Typical case

Case one

A 71-year-old woman presented with lumbar and back pain following trauma one year prior. Hospital admission occurred due to worsening pain over the preceding ten days. Preoperative radiographs (A, B) demonstrated wedge deformity of the L1 vertebral body. Preoperative lumbar MRI (T1, T2 sequences) (C, E) revealed collapse of the superior endplate of the L1 vertebral body, with low signal intensity within the vertebral body accompanied by vacuum fissure syndrome. Fat-suppressed MRI (D) showed: high signal intensity within the vertebral canal due to oedema and fluid accumulation (Fig. 11). Clinical diagnosis: Non-union of an old L1 vertebral fracture. Navigation-guided percutaneous vertebroplasty (PVP) was performed. (A, B) depict intraoperative navigation imaging; (C, D) show intraoperative fluoroscopic images (Fig. 12). Postoperative radiographs (A, B) demonstrated: Good cement filling of the L1 vertebral body. Postoperative three-dimensional CT (C-F) showed: Diffuse cement distribution with no evidence of leakage (Fig. 13).

Fig. 11.

Preoperative lumbar spine X-ray and MRI examinations (A) (B): Preoperative radiographs; (C) (E): Preoperative lumbar MRI (T1, T2 sequences); (D) Fat-suppressed MRI

Fig. 12.

Intraoperative navigation real-time imaging, intraoperative fluoroscopic imaging (A) (B): Intraoperative navigation imaging; (C) (D): Intraoperative fluoroscopic images

Fig. 13.

Postoperative re-examination: lumbar spine X-ray and 3D CT scan (A) (B): Postoperative radiographs; (C) (D) (E) (F): Postoperative three-dimensional CT

Case two

An 81-year-old female presented with lumbar and back pain following bending to lift heavy objects three days prior. Preoperative radiographs (A, B) revealed wedge deformity of the L2 vertebral body. Preoperative lumbar MRI (T1, T2 sequences) (C, D) demonstrated collapse of the inferior endplate of the L2 vertebral body with low signal intensity within the vertebral body. Fat-suppressed MRI (E) demonstrated: high signal intensity within the vertebral canal due to oedema (Fig. 14). Clinical diagnosis: fresh compression fracture of the L2 vertebral body. Navigation-assisted percutaneous vertebroplasty (PVP) was performed. (A, B) show intraoperative navigation images; (C, D) show intraoperative fluoroscopic imaging (Fig. 15). Postoperative radiographs (A, B) indicate: Good cement filling of the L2 vertebral body. Postoperative three-dimensional CT (C-F) indicates: Uniform cement distribution with good dispersion and no leakage (Fig. 16).

Fig. 14.

Preoperative lumbar spine X-ray and MRI examinations (A) (B): Preoperative radiographs; (C) (D): Preoperative lumbar MRI (T1, T2 sequences); (E) Fat-suppressed MRI

Fig. 15.

Intraoperative navigation real-time imaging, intraoperative fluoroscopic imaging (A) (B): Intraoperative navigation imaging; (C) (D): Intraoperative fluoroscopic images

Fig. 16.

Postoperative re-examination: lumbar spine X-ray and 3D CT scan (A) (B): Postoperative radiographs; (C) (D) (E) (F): Postoperative three-dimensional CT 

Discussion

This prospective randomised controlled study aimed to determine the clinical efficacy of the computer-guided percutaneous vertebroplasty in treating lumbar osteoporotic vertebral fractures and to analyse radiographic outcomes. The findings indicate that the navigation-assisted PVP procedures demonstrate clinical feasibility and yield favourable clinical outcomes. This approach enhances puncture accuracy and surgical safety, reduces the number and time of punctures, minimises intraoperative radiation exposure for both medical staff and patients, and lowers the incidence of complications such as bone cement leakage. It achieves ‘visualisation’ of surgical procedures including puncture, thereby better realising ‘precision medicine’. The Visual Analogue Scale (VAS) and the Oswestry Disability Index (ODI) are crucial indicators for assessing patient pain levels and quality of life. Consequently, they are commonly employed to evaluate clinical efficacy before and after treatment [22]. In this study, we compared the clinical efficacy between the navigation group and the traditional group using VAS and ODI scores. Our findings demonstrated that postoperative VAS pain scores were significantly reduced in both groups compared to preoperative levels, with marked improvements in quality of life. Previous studies suggest that the thermal effect of bone cement causes degeneration and necrosis of neural elements at the vertebral ends. Postoperatively, the injured vertebrae experience reduced compression and fracture lines are stabilised by bone cement adhesion, leading to marked pain relief in the short term [23]. The significant changes in VAS and ODI scores before and after surgery indicate that the application of the navigation systems in PVP techniques constitutes an effective adjunctive method. During routine postoperative follow-up, both groups demonstrated improved VAS and ODI scores at various time points compared to preoperative levels, with no statistically significant differences between groups (P > 0.05). This indicates that the computer-navigated PVP also achieves favourable long-term therapeutic maintenance, consistent with findings by Li et al. [24]. Postoperative HFV improved in both groups compared to preoperative levels, suggesting that the computer-navigated PVP also effectively restores vertebral height. Furthermore, during the comparative analysis, we observed that the computer-navigated PVP technique yielded lower pain feedback (VAS scores) during intraoperative puncture to the target site compared to the traditional group (P < 0.05). This may be attributed to reduced intraoperative discomfort caused by repeated needle repositioning and fluoroscopic adjustments, along with diminished stimulation of periosteal surfaces and muscles, and lessened soft tissue injury [25], thereby enhancing the patient’s procedural experience.

The computer-guided navigation systems significantly reduce the number of punctures and puncture time, with statistically significant differences between groups (P < 0.001). This is because, following two- and three-dimensional image registration, precise punctures can be performed in real time via the central console along the preoperative planning pathway, enhancing surgical safety and accuracy. Under visual guidance, punctures can be directed to the fracture site and establish an optimal bone cement infusion pathway, thereby substantially reducing the number of punctures and puncture time ([26]. Although computer-navigated PVP technology substantially reduces puncture counts, the initial implementation phase involved prolonged pre-puncture registration times, with the longest session reaching 28 min, resulting in total surgical times exceeding those of the conventional fluoroscopy group. As anticipated, total procedure times are influenced by the learning curve associated with navigation system utilisation ([27]. With increased procedural experience, enhanced proficiency in navigation equipment operation, improved coordination between surgeons and technicians, and a shortened learning curve, registration time markedly decreased compared to initial implementation. In subsequent applications, the total surgical time for the navigation group even fell below that of the conventional fluoroscopy group. This aligns with findings by Yuan et al. ^[7], indicating that total surgical time in the navigation group progressively decreases with increasing case numbers. However, no statistically significant difference in total procedure duration was observed between the two groups in this study (P > 0.05).

Traditional fluoroscopy-guided percutaneous vertebral puncture requires repeated fluoroscopic adjustments to refine needle positioning, ensuring both accuracy and safety. Previous studies indicate that minimally invasive spinal surgery under C-arm fluoroscopy significantly increases spinal surgeons’ radiation exposure through multiple fluoroscopic sessions, potentially leading to serious consequences such as cataracts, skin inflammation, and cancer [28–30]. Navigation-assisted PVP technology requires only a single pre-puncture C-arm fluoroscopy in an anteroposterior and lateral view for two- and three-dimensional image registration. During the procedure, two additional anteroposterior and lateral fluoroscopies verify the safety and accuracy of the anatomical positioning images. Consequently, the computer navigation systems for PVP significantly reduce the total number of fluoroscopy views and the overall fluoroscopy time, thereby decreasing radiation exposure for both medical staff and patients. A prospective case-control study by Hu et al. [31] involving 18 patients demonstrated that computer-assisted spinal surgery systems are clinically applicable for PVP. Utilising such systems to guide procedures provides more precise entry points and reduces unnecessary radiation exposure. Strong et al. [32] confirmed that computer-assisted 3D navigation in treating adult spinal deformities reduces surgeons’ fluoroscopic radiation exposure, consistent with the findings of this study.

In conventional surgical approaches, achieving a satisfactory puncture pathway necessitates repeated intraoperative adjustments to the needle insertion site and its head-tail tilt and lateral inclination angles. This may compromise the medial and lateral walls of the pedicle and exacerbate damage to osteoporotic vertebral bodies, thereby increasing the risk of bone cement leakage and associated complications. Reports indicate that bone cement leakage rates during vertebroplasty currently range from 23% to 62% [33–35]. Leakage into the spinal canal or intervertebral foramen may cause compression of lower limb nerve roots, leading to incomplete or complete paralysis [36]; while leakage into arteries may pose severe risks including vascular embolism and associated complications [37, 38]. In this study, neither group experienced major vascular injury, nerve injury, or vascular embolism. Cement leakage occurred in 2 cases (6.25%) in the navigation group, while the traditional group experienced cement leakage in 9 cases (30.00%). The incidence of cement leakage in the navigation group was lower than that in the traditional group, with a statistically significant difference (P < 0.05). All cement leakage cases occurred within the intervertebral disc space or the anterior/posterior/lateral aspects of the vertebral body. Cement leakage in both groups did not cause neurological symptoms or discomfort in patients. The application of the navigation system to assist PVP enables preoperative planning of the puncture pathway. During surgery, the real-time position of the optical tracking ball is captured via an infrared camera. Subsequently, a single anteroposterior and lateral fluoroscopic image is acquired. This intraoperative fluoroscopic image is then matched with the preoperatively planned vertebral image. Upon successful matching, the position of the puncture needle is determined in real time, rendering the entire puncture procedure ‘visualised’. Subsequently, a single puncture accurately reached the pre-planned target site along the pre-planned pathway, enabling precise and rapid injection of bone cement into the vertebral target location. The navigation group utilised a greater volume of bone cement (6.19 ± 1.37 ml) compared to the traditional group (5.45 ± 1.02 ml), with a statistically significant difference between the two groups (P < 0.05). This suggests that the navigation-assisted puncture enables more precise targeting of the fracture gap, thereby facilitating greater bone cement filling. Navigation-assisted PVP procedures not only mitigate the risk of further vertebral damage and cement leakage caused by repeated needle repositioning but also facilitate superior cement dispersion and distribution along the fracture site, thereby achieving greater cement utilisation. These findings align with those of a meta-analysis [39].

Computer-assisted navigation systems employ spatial three-dimensional real-time navigation technology and computer-aided image processing techniques to process and integrate preoperative or intraoperative imaging data. They perform three-dimensional reconstruction of the patient’s actual position and anatomical structures, projecting navigation tools and surgical instruments onto virtual computer-simulated animated images in real time. This enables surgeons to continuously monitor the precise relative positioning of surgical instruments against anatomical landmarks based on the provided navigation data. Through seamless integration from preoperative planning to intraoperative execution, to achieve greater visualisation of the surgical procedure. However, during the early stages of this technology’s implementation, to achieve rigid fixation of the patient reference frame to the patient’s vertebral bodies, two to three K-wires were required to fix the spinous process locator onto the spinous processes of the vertebrae one or two levels below the affected segment [40]. This ensured the absolute bony spatial relationship between the locator, the connected patient reference frame, and the spine, minimising pairing errors and subsequent puncture inaccuracies caused by relative displacement or respiratory movement. Nevertheless, this approach increased patient trauma and discomfort. In this study, we attempted to secure the spinous process locator using skin adhesive tape. During the procedure, we observed that this method occasionally resulted in inaccurate matching or positional shifts, ultimately leading to puncture failure or increased overall puncture and surgical times. With accumulated experience, improved procedural proficiency, and refinements to fixation techniques, we subsequently achieved favourable outcomes by employing skin adhesive tape to securely affix the patient reference frame to the central position of the vertebral body one or two levels below the fracture. Tao Hui et al. [41] demonstrated that an improved tracer fixation technique (i.e., skin tape fixation) enables safe and effective surgical completion with favourable clinical outcomes; it reduces trauma and facilitates earlier patient recovery. Naturally, we are also designing novel fixation methods to achieve more stable relative positioning while minimising additional patient trauma and discomfort.

This computer navigation system employs infrared cameras (NDI) to capture optical tracking spheres mounted on the patient reference frame, image register, and needle stabiliser for real-time positional tracking. Consequently, no obstructions may exist between these components, necessitating absolute transparency. Any obstruction may result in loss of spatial positioning, leading to failed matching and ultimately causing significant errors that could precipitate puncture failure or elevate surgical risks. This finding aligns with our research experience and previous studies [42], confirming that the optical tracking spheres on the patient reference frame, image registrar, and needle stabiliser must remain unobstructed during surgery.

Rahmathulla et al. [43] identified potential challenges in achieving adequate imaging for obese and morbidly obese patients. Increased soft tissue in morbidly obese patients complicates patient positioning, beam penetration, and the ability to maneuver imaging equipment around the patient. This results in poor image quality, leading to inaccurate registration and rendering images difficult to utilize during surgery. In this study, we observed that for severely obese patients (BMI ≥ 30 kg/m²), the excessive thickness of the fat layer at the lumbar fracture site not only resulted in poor intraoperative fluoroscopic image quality but also made it difficult to achieve a relatively fixed position for the patient reference frame and the fractured vertebral body, with a significant distance between them. This led to difficulties in obtaining spatial positions during the initial phase and subsequent changes in relative positions causing unsuccessful matching, ultimately resulting in puncture failure and increased surgical risk. During the puncture process, excessively thick fat may prevent the needle from reaching the preoperative target location or cause errors or loss of real-time needle tip tracking when the needle is positioned at a deeper depth.

During the initial stages of our research, we observed that excessive respiratory movement in patients adversely affects intraoperative two- and three-dimensional alignment, potentially leading to failed needle placement. The NDI primarily captures and localises the positions of optical tracking spheres on the patient reference frame and image registrator. It converts these two-dimensional images into stereoscopic three-dimensional forms via anteroposterior and lateral fluoroscopy. Consequently, excessive respiratory motion during fluoroscopy may cause discrepancies between the fluoroscopic image position and the NDI’s recognised and captured positions, potentially resulting in registration failure or even puncture failure. Huang et al. [44] observed that both respiratory amplitude and movement induced by the puncture procedure may compromise surgical navigation system accuracy. This aligns with our findings. The lumbar spine exhibits smaller respiratory-induced displacement relative to the thoracic spine, constituting a key rationale for selecting the lumbar region as the primary focus of our study.

This computer-assisted navigation technology offers the following advantages in lumbar percutaneous vertebroplasty (PVP) applications: (1)Preoperative needle path planning enhances intraoperative puncture accuracy and surgical safety, reducing puncture time and the incidence of procedure-related complications; (2)Requires only one pre-puncture C-arm fluoroscopy in anteroposterior and lateral views; during the procedure, anatomical positioning is verified via two additional anteroposterior and lateral fluoroscopies, significantly reducing intraoperative fluoroscopy frequency and radiation exposure; (3)Intraoperative puncture follows the pre-planned pathway, rendering the entire procedure ‘visualisable’. The needle position can be ‘visualised’ and adjusted in real-time within the navigation imaging system according to the surgeon’s personal technique; (4)Compared with O-arm navigation, intraoperative CT navigation, and robot-assisted PVP technology, this computer navigation technology has the advantages of being easy to operate, having lower costs, and not requiring intraoperative scanning of 3D CT, which would otherwise increase the radiation dose for both medical staff and patients [45, 46]; (5)This assistive technology serves as a valuable aid for junior clinicians, maybe help to shorten the learning curve for PVP techniques.

Limitations

This study also has certain limitations: firstly, the sample size is relatively small, potentially introducing statistical bias. Future research should incorporate more cases to highlight further advantages of computer-assisted navigation in descending PVP surgery. Secondly, the follow-up period was relatively short, with assessments conducted only at 3, 6, and 12 months. Consequently, the long-term therapeutic outcomes and complications remain incompletely observed, and statistical analysis of radiation-related diseases among medical personnel could not be performed. Finally, this study excluded patients with degenerative scoliosis complicated by vertebral rotation (Cobb angle > 25°), those with severe obesity (BMI ≥ 30 kg/m²), and those with thoracic vertebral compression fracture. These conditions posed distinct challenges that could lead to puncture failure or increased surgical risk, thereby limiting the study population. In future endeavours, with the rapid advancement of artificial intelligence and the swift progress of software and hardware technologies, we shall pursue improvements across all aspects. We are confident that this computer-assisted navigation technology will benefit a broader spectrum of patients with spinal disorders in the years to come.

Conclusion

This computer-assisted navigation system enhances the precision of PVP puncture and surgical safety, increases the volume of bone cement used by precisely and effectively filling the fracture gap, while reducing the frequency and duration of fluoroscopy. It minimises radiation exposure for both patients and medical staff during operations, decreases the number of punctures required, thereby improving the patient’s experience. By enabling visualisation of surgical procedures, it facilitates precision medicine.

Authors’ contributions

*YI LIAN and YANCHUN XIE contributed equally to this study. They are responsible for data processing and analysis, work summarisation, and the final drafting of articles.HAILONG YU is responsible for final review and proofreading.ANWU XUAN, YONGCUN WEI and ZENING WANG were responsible for reviewing the data and references.LIANGBI XIANG and HONGWEN GU were responsible for image production.All authors reviewed the manuscript.

Funding

No funding was received.

Data availability

The data underpinning the conclusions of this study may be obtained from the General Hospital of Northern Theater Command. Owing to the hospital’s specialised nature, patient data is subject to access restrictions—its use requires authorisation and is not publicly available. However, upon reasonable request and with permission from the General Hospital of Northern Theater Command, the relevant data may be obtained from the authors.

Declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committee of General Hospital of Northern Theater Command. Ethics approval number: Lun Shen Y (2021) No. 025. and informed consent was obtained from all patients.The study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki.

Consent for publication

Not applicable

Competing interests

The authors declare no competing interests.