Application of Augmented Reality for Accurate Punctures During Stage 1 Sacral Neuromodulation

Article information

Int Neurourol J. 2024;28(4):302-311

Publication date (electronic) : 2024 December 31

doi : https://doi.org/10.5213/inj.2448330.165

Haoyu Yuan^*, Yuansong Xiao^*, Xiaoyu Lin, Lei Zhang, Chenglin Yang, Zhengfei Hu, Yue Yang, Peixian Wu

Department of Urology, General Hospital of Southern Theater Command, Guangzhou, China

Corresponding author:Yuansong Xiao Department of Urology, General Hospital of Southern Theater Command, 111 Liuhua Road, Liuhua Street, Yuexiu District, Guangzhou, 510010, China, Email: yuansongdoctor@163.com

*Haoyu Yuan and Yuansong Xiao contributed equally to this study as co-first authors.

Received 2024 August 18; Accepted 2024 December 3.

Abstract

INTRODUCTION

Voiding dysfunction, a prevalent condition in urology, is characterized by lower urinary tract dysfunction stemming from various causes. It encompasses several subcategories, including lower urinary tract symptoms, overactive bladder (OAB), nonobstructive urinary retention (NOR), neurogenic lower urinary tract dysfunction (NLUTD), and interstitial cystitis/bladder pain syndrome (IC/BPS) [1]. The initial treatment options typically include drug therapy and pelvic floor functional exercises. However, these conservative approaches often have limited effectiveness for patients with intractable voiding dysfunction, necessitating further third-line treatments such as sacral nerve modulation and botulinum toxin injections [2, 3].

Synergistic micturition can be achieved by stimulating the lower central sacral nerve, which controls urination [4, 5]. After clinical verification, the U.S. Food and Drug Administration approved sacral neuromodulation (SNM) for the treatment of urgent urinary incontinence in 1997. The indications were expanded in 1999 to include frequency-urgency syndrome, nonobstructive urinary retention, and defecation dysfunction [6].

SNM is divided into 2 stages: The success of implanting the stage II stimulator is significantly influenced by the type of disease and the accurate placement of the SNM electrode during stage I. Traditionally, clinicians use x-ray fluoroscopy to identify the number and size of the sacral foramen, the shape of the sacrum, the sacroiliac joint, and other anatomical features prior to surgery. During the procedure, they test the motor and sensory responses of the patients to verify the accuracy of the needle insertion site. However, the quality of these images can be compromised by the presence of pelvic organs, intestinal contents, and other factors, which may obscure bone changes. For patients with deformities of the sacrum and coccyx, enhanced pelvic computed tomography (CT) can provide a better understanding of pelvic anatomy and help identify sacral variations. Nevertheless, it remains challenging to gain a comprehensive understanding of the 3-dimensional anatomy of the sacral nerve foramina [7]. Therefore, it is necessary to develop a more convenient and efficient localization method to improve the accuracy of sacral foramen punctures. Currently, there are studies exploring technologies such as 3-dimensional (3D) printing and ultrasound guidance [8–10]. However, their effectiveness and safety still require further evaluation.

Augmented reality (AR) is an emerging technology that creates simulated virtual objects and scenes, which are then superimposed onto the real world. AR can provide surgeons with 3D anatomical information about the sacral foramen and sacral nerve during puncture procedures, serving as an auxiliary tool during operations. There is already evidence supporting the use of AR in surgeries for prostate cancer, kidney cancer, bladder cancer, and lithiasis [11, 12]. However, its application in sacral nerve regulation, along with its advantages and disadvantages, requires further investigation [9, 13].

MATERIALS AND METHODS

Materials

Objects

This retrospective study analyzed the clinical data of 130 patients who underwent sacral nerve regulatory electrode implantation in our department from 2018 to 2023 (Fig. 1). These patients received comprehensive care—from preoperative preparation through surgical operation to postoperative follow-up management—under the guidance of the same senior professional physician.

Fig. 1

Flowchart of the study. OAB, overactive bladder; NOR, nonobstructive urinary retention; NLUTD, neurogenic lower urinary tract dysfunction; IC/BPS, interstitial cystitis/bladder pain syndrome; SNM, sacral neuromodulation; AR, augmented reality.

Criteria

The inclusion criteria were as follows:

1. (1) Indications recommended by clinical experts in OAB, NOR, NLUTD, IC/BPS

2. (2) Patients older than 16 years old who were physically well-developed

3. (3) The operation performed by the same operator

The exclusion criteria were:

1. (1) An uncontrollable complex urinary tract infection

2. (2) Severe impairment of upper urinary tract function (severe vesicoureteral reflux, severe hydronephrosis, etc.)

3. (3) Bladder outlet obstruction

4. (4) Minors with progressive neuropathy, complete paraplegia, or congenital neurological disorders

5. (5) Mentally handicapped persons who could not use the device or evaluate its efficacy

6. (6) Pregnant women

Experimental equipment

The following equipment was used: a mobile C-arm x-ray machine (Siemens, Munich, Germany), a 256-row spiral CT detector (Philips, Amsterdam, The Netherlands), sacral nerve stimulator assembly (Medtronic, Dublin, Ireland), and HoloLens2 holographic glasses (Microsoft, Redmond, WA, USA).

Research Methods

Clinical information collection and grouping of patients

The medical records of both outpatients and inpatients were retrospectively collected. These records included, but were not limited to, the primary diagnosis (determined after a urodynamic examination), symptoms, disease progression, treatment history, past medical history, and any neurosurgical interventions. Additionally, laboratory and imaging examinations conducted after admission were reviewed, along with specialist evaluations such as the 72-hour micturition diary and further urodynamic testing. Data regarding the surgical procedure were also gathered from the operation records. This included the duration of the operation, volume of blood loss, whether there was a need for additional anesthesia injections, number of punctures, number of electrode contact responses, initial response voltage, frequency of x-ray use, and others. Furthermore, the postoperative pain was assessed using a numeric rating scale.

1) Implantation of sacral nerve regulatory electrodes under traditional x-ray localization

The standard treatment for SNM involves a 2-stage process. Initially, a sacral nerve regulatory electrode is implanted, followed by an observation period of 2–4 weeks. Subsequently, a permanent sacral nerve stimulator is implanted. The first stage of the procedure primarily concentrates on accurately locating the sacral foramen.

2) Implantation of sacral nerve regulation electrode guided by hologram

Holographic AR guidance facilitates the localization and puncture of the sacral foramen during the electrode implantation in sacral neuroregulatory devices.

Preoperative preparation: A plain CT scan was taken of the lumbar vertebrae and sacrococcyx before the operation, with the following steps:

1. Posture: the patient was in a prone position with a high lower abdomen so that the posterior surface of the sacrum was in a horizontal position. This position was kept consistent with the intraoperative posture (Fig. 2A).

2. Determination of positioning points: We defined the first positioning point as being located in the middle nodding side of the posterior superior iliac spine line. The second and third positioning points were situated 9 cm cephalic from the apex of the coccyx and 4 cm laterally to the left and right, respectively. These 3 points were triangularly distributed within the puncture area. Care was taken to ensure that they were not occluded during the operation (Fig. 2A).

3. CT scan of the lumbosacral region: A 1-mm thin-slice CT scan was performed, encompassing at least the fifth lumbar vertebra on the cephalic side and the ischial tubercle on the caudal side.

4. CT data processing: We collaborated with professional engineers to import DICOM data from plain CT scans into the Mimics 3D reconstruction system. This process generated a 3D mesh model (Fig. 2B). Subsequently, the 3D data were transmitted over the network to Microsoft HoloLens2 holographic glasses.

Fig. 2

(A) The selection of anchor points. Three points are triangularly distributed in the puncture area, with care taken to ensure that the 3 positioning points are not occluded. (B) Three-dimensional reconstruction images were obtained based on a plain computed tomography (CT) scan, and the puncture path was designed. The bones, nerves, blood vessels, and muscles in the target area were labeled according to different CT values. After visual detection, segmentation, extraction, measurement, verification, and calculation, a 3-dimensional mesh model is generated.

The intraoperative operation was as follows:

1. The patient’s surgical position was consistent with the position used in the CT examination.

2. Fusion of hologram and real-time picture: Prior to disinfection, the operator wore HoloLens2 glasses, projecting the AR equipment’s hologram onto the patient’s body surface. The hologram was aligned with the body surface marks and positioning points, allowing for precise marking of the bilateral skin puncture needle points based on the holographic measurement data.

3. Routine disinfection of the operation area and placement of an aseptic sheet, taking care to avoid covering the fixed point (Fig. 3A): The surgeon, wearing HoloLens2 glasses, reconfirmed the alignment of the hologram with the body surface marks. Holding the puncture needle, the surgeon then adjusted it to match the position and multiazimuth angle depicted in the hologram in real-time (Fig. 3B).

4. C-arm fluoroscopy to check whether the position of the puncture needle was satisfactory and make slight adjustments if needed.

5. Testing of the motor response and sensory response of the patients to further determine whether the puncture site was correct (Fig. 3C).

Fig. 3

(A) The body surface mark of the location point was maintained until before the operation. (B) The puncture and real-time fusion picture under the guidance of augmented reality during operation. (C) The fusion picture shown on augmented reality glasses and x-ray confirmed the electrode position again after the electrode was successfully implanted.

During the operation, a successful outcome was indicated by the sensory response of the perineum, movement of the perineal bellows, and metatarsal flexion of the big toe. This was assessed in conjunction with the positioning of an x-ray electrode at the S2, S3, or S4 locations.

RESULTS

Demographic and Clinical Characteristics of Patients

This study reviewed 130 patients who underwent SNM electrode implantation in our department from January 2018 to December 2023. The test group underwent puncture guided by AR, while the control group received puncture under conventional x-ray localization. There were no significant differences between the 2 groups in terms of sex, age, main diagnosis following urodynamic examination (OAB, NOR, NLUTD, IC/BPS), underlying diseases, and routine preoperative examination items after admission, among other general data (Table 1). However, there was a significant difference in the history of lumbosacral-related diseases between the 2 groups (P<0.001), which is discussed in subsequent sections.

Table 1

Comparisons of clinical characteristics between experiment group and control group (n=130)

Comparison of Intraoperative Conditions of the Sacral Nerve Regulatory Electrode Implantation Test

In the experimental group, the average number of punctures was 4.29, whereas the conventional positioning method required approximately 6 adjustments per puncture to meet the needs of electrode implantation surgery. The data also indicate that AR-guided positioning punctures tend to be more accurate. Under AR guidance, the puncture needle is positioned closer to the nerve, resulting in a higher number of responding electrodes (2.74±0.51 vs. 2.46±0.74) and a lower initial voltage postimplantation (1.09±0.39 V vs. 1.69±0.43 V). Additionally, the use of repeated x-ray fluoroscopy for adjustments and confirmations during the puncture process was significantly reduced in the test group compared to the control group (5.94± 1.46 vs. 9.22±1.93). Consequently, the overall operation time was shorter in the test group (61.32±11.27 minutes vs. 83.49± 15.84 minutes). There was no significant difference between the 2 groups in terms of the amount of blood loss and the positioning of the sacral foramen implants (Table 2).

Table 2

Comparison of intraoperative conditions of sacral neuromodulation electrode implantation-testing (n=130)

Comparison of the Curative Effect of Implantation of Sacral Nerve Regulatory Electrodes

The rate of permanent implantation was significantly higher in the experimental group compared to the control group (76.47% vs. 57.29%). Additionally, the pain score on the first day after the operation was significantly lower in the experimental group than in the control group (2.62±0.697 vs. 2.83±0.816). This may be attributed to fewer puncture attempts and a shorter operation duration. There was no significant difference in the incidence of wound infection between the 2 groups (Table 3).

Table 3

Comparison of the curative effect of implantation of sacral neuromodulation electrode

DISCUSSION

Voiding dysfunction encompasses a range of urological diseases, with drug therapy and pelvic floor function exercises serving as the primary treatment options. However, these conservative treatments often have limited efficacy for patients with refractory urinary dysfunction, leading to the need for further third-line surgical interventions. Precise electrode placement is essential for the success of SNM. SNM improves nerve conduction in both neurogenic and nonneurogenic diseases by employing nerve electrical stimulation and neuromodulation [5, 14]. In this technique, the electrode is implanted through interventional methods, and a short pulse stimulation current is continuously applied to the sacral nerve. This stimulation influences the local sacral nerve reflex and the feedback response from the higher center via the afferent nerve pathway. The process of “neural regulation” is achieved by stimulating and regulating the activity of sacral nerve effector organs such as the bladder, urethral sphincter, and pelvic floor muscle. Increasing clinical studies and applications [15, 16] indicate that SNM can significantly improve symptoms in patients with pelvic floor disorders, including urgent urinary incontinence, refractory bladder overactivity, nonobstructive urinary retention, defecation dysfunction, and neurogenic bladder. However, the precise mechanism of SNM still requires further investigation.

The normal innervation of the lower urinary tract involves 3 major nerve categories: sympathetic, parasympathetic, and somatic nerves [14]. Postganglionic fibers from the T11–L2 segment act on the bladder plexus to relax the smooth muscle of the bladder wall and contract the smooth muscle of the bladder triangle, the bladder sphincter, and the urethra near the bladder, facilitating urine storage. Parasympathetic preganglionic fibers originating from the S2–4 segment travel via the pelvic nerve to the bladder plexus, distributing to the bladder wall to control its contraction and promote bladder emptying. The somatic nerve, originating from the pudendal nerve—which is composed of the anterior root of the sacral nerve from S2–4—has an afferent component that receives sensations from the prostatic part of the urethra. Its efferent component transmits sensations from the external urethral sphincter to the cavernous muscle of the urethra. This efferent nerve also aids in contracting the internal sphincter of the urethra and closing the bladder outlet to retain urine. Most innervation of the bladder and urethra is transmitted through the S2–4 nerve segments via the pelvic nerve, which is a target for treating voiding dysfunction through electrical stimulation [17].

According to clinical research, experts agree that S3 is the preferred target for SNM punctures [3, 18]. The ideal position for electrode placement is defined by the following criteria: 1. The starting voltage of all contacts does not exceed 2V; 2. During the gradual increase in stimulation intensity, there should be a perineal sensory response, perineal bellows movement, and metatarsal flexion movement of the big toe. Intraoperative sensory and motor responses are crucial for the success of the operation. The S3 sacral foramen localization method under x-ray fluoroscopy is employed in standard sacral nerve regulatory electrode implantation [19]. During the operation, it was observed that repeated x-ray fluoroscopy was necessary to confirm the projection point of the sacral foramen on the body surface and to adjust the position of the puncture needle and electrode based on the sensory and motor responses. This was particularly true for patients with a history of lumbosacral trauma or congenital sacral deformities, where the variant anatomical structure significantly increased the duration of the operation and the number of x-ray fluoroscopies. Consequently, there was an increase in the likelihood of infection and the intensity of pain experienced by the patients. Additionally, the radiation dose received by both patients and operating room personnel was substantially higher. Therefore, it is essential to explore a more accurate method for positioning the puncture.

Structural abnormalities in the lumbosacral region are considered a relative contraindication for the implantation of sacral nerve regulatory electrodes, as noted in some studies [19, 20]. This is primarily because variations in bone and nerve structure significantly complicate the processes of locating the correct site and performing the puncture. Additionally, it may be challenging to elicit an effective electrode shock response during surgery, or there may be a need to start with an initial voltage higher than the 2 V recommended by the guidelines. We assessed the structural variations of the lumbosacral region based on the patient’s medical history, physical examination at admission, and routine imaging results. Priority was given to reconstructing a 3-dimensional model of the surgical area and utilizing AR guidance during the procedure. Among the 130 patients evaluated, 28 exhibited abnormal bone structures of the lumbar vertebrae or sacrococcyx, or nerve deformities. These abnormalities primarily included changes following lumbosacral trauma, sacral dysplasia, lumbosacral transitional vertebrae, asymmetry of the sacral articular processes, and anterior sacral dural protrusion. The incidence of sacral deformity was higher in the experimental group compared to the control group. Based on patient preferences, 15 patients (53.57%) underwent AR-guided SNM electrode implantation, while 13 patients (46.43%) had the procedure guided by routine x-ray localization. In the group with normal lumbosacral structures, 19 patients opted for AR-guided surgery. We speculate that these choices contribute to the statistically significant differences observed between the test group and the control group.

The more accurately the electrode aligns with the target sacral nerve, the greater the number of contacts that respond during testing. This alignment allows for satisfactory sensory and motor responses at lower starting voltages, thereby extending the battery life [10, 21]. In our study, we discovered that the use of holographic technology in the implantation of sacral nerve-controlled devices not only increases the number of effective electrode contacts but also reduces the initial voltage required. This suggests that the electrode placement is more precisely near the nerve. We hypothesize that this will also enhance the stability and longevity of the device postoperatively. The use of intraoperative hologram technology significantly reduces the need for x-ray fluoroscopies, thereby considerably lowering both the frequency of radiation exposure and the radiation dose for operating room staff who frequently perform these procedures. This approach underscores an important principle of patient-centered medicine. In a retrospective study, 1 patient who underwent sacral nerve modulation developed a wound infection during home observation postsurgery. The infection was caused by Staphylococcus aureus, and the persistent wound infection ultimately led to the removal of the implant. It is crucial to manage perioperative wound care effectively and to emphasize the importance of wound care education for patients monitored at home following surgery [22].

We believe that AR guidance offers several advantages in the implantation of sacral nerve control electrodes: (1) The puncture angle is more accurate, as evidenced by increased electrode reaction contacts, lower initial voltage, and reduced postoperative pain scores, which may also decrease the learning curve. (2) Reduced x-ray exposure due to fewer intraoperative fluoroscopies. (3) Better suited for patients with variations and deformities of the sacral vertebrae. (4) More convenient intraoperative adjustments, characterized by fewer electrode adjustments and reduced operation time. (5) The low learning threshold and convenience of head-mounted AR devices can shorten the learning curve for junior doctors performing stage 1 SNM compared to traditional methods. However, there are also limitations: (1) Dependence on AR devices and professional engineers makes the technology challenging to implement at the grassroots level. (2) The additional cost may not be acceptable to every patient. (3) Necessary location and preparation before the operation may prolong the hospital stay. The limitations of this paper include: (1) It is a retrospective study, lacking a prospective cohort study. (2) It is a single-center study with an insufficient sample size. (3) We did not compare various localization methods, such as ultrasonic localization, 3D printing guide boards, and cognitive fusion puncture. (4) Further studies are needed for the systematic classification of patients with anatomical abnormalities. These areas will also be the focus of our future research.

In this study, we retrospectively collected data from 130 patients who underwent SNM electrode implantation in our department. These patients were categorized based on whether they received holographic guidance during the procedure. We hypothesize that AR-guided SNM electrode implantation can decrease the number of punctures and the duration of the operation while enhancing the precision of electrode placement, all without compromising safety and effectiveness. Furthermore, increasing the number of response electrode contacts appears to effectively lower the initial response voltage, thereby stabilizing the electrode. This technique could serve as a valuable adjunct technology to enhance the accuracy of SNM electrode implantation, particularly in patients with abnormalities in sacrococcygeal skeletal nerve anatomy.

Article information Continued