Introduction

An aneurysm is characterized by an abnormal widening of the vessel lumen exceeding 50% of the normal size, caused by weakened or abnormal spots of the vessel wall, affecting all three vessel layers [1, 2]. Intracranial subtype specifically affects the cerebral arteries [3]. Treatment modalities include clip**, coiling, flow diverter (FD), stent placement, or artery sacrifice in cases where traditional methods are not applicable.

Integrating novel devices into surgical procedures proves invaluable in addressing these challenges. The evolution of surgical robots, starting with the PUMA 200 in 1985, has led to the widespread adoption of advanced robotic surgical devices in operating theatres globally [4].

The demand for precision in procedures, especially in the intricate realm of intracranial aneurysms, provides an ideal platform for the application and evaluation of surgical robots. The potential repercussions of even a minor mistake, such as the rupture of an arterial wall leading to haemorrhage or patient fatality, underscore the critical need for precise interventions. Operating robots have previously demonstrated success in various procedures, notably in aortic vascular operations, including the treatment of abdominal aortic aneurysms, the most prevalent type of aneurysm [5, 6]. Encouraged by these outcomes, we sought to explore the effectiveness of utilizing operating robots in the treatment of cerebral aneurysms.

Currently, there is a notable absence of a systematic review that effectively consolidates knowledge exclusively related to this subject. Our efforts in this study are concentrated on examining clinical applications of robotic devices in surgical interventions for intracranial aneurysms in humans. We specifically emphasize treatment applications rather than diagnosis, excluding angiography from our scope.

Recognizing the significance of consolidating existing knowledge, our aim is to contribute to the understanding of how the implementation of these robotic devices is likely to reshape the work routines of neurosurgeons in the near future. Furthermore, our aim is to address the current limitations of robotic systems for future research in this field.

Methodology

We adhered to the PRISMA guidelines for reporting systematic reviews [7]. The literature search process involved seven major medical databases: PubMed, Scopus, Embase, Web of Science, Cochrane Library, Ebsco Host, and Ovid. MeSH terms, including “cerebral aneurysm,” “brain aneurysm,” “intracranial aneurysm,” and “robot” or “robot-assisted” or “guide robot,” were employed in the search. The full search strategy is available in Appendix 1. No search filters or specific publication date ranges were applied.

Inclusion criteria encompassed (I) robot-assisted cerebral aneurysm intervention procedures with (II) full-text availability as an original study or case report. Exclusion criteria covered (III) in vitro studies (e.g., phantoms) or (IV) non-human in-vivo studies (such as porcine models).While phantom/ex-vivo testing is valuable for evaluating new technologies, our focus was on clinical outcomes and practical effectiveness of this technology. Non-human studies, primarily aimed at develo** innovations, cannot fully assess these aspects, compared to in-human interventions. Additionally, (V) abstract-only works, (VI) review articles, technical descriptions, and other non-original articles, (VII) reports focused on diagnostics (angiography, microscope assist), and (VIII) works lacking full-text access were excluded. Articles not in English or Polish were translated using Google Translate and DeepL translators.

The search process was independently conducted by P.Ł. & B.J., with conflicts resolved by the third author K.J.

Data extraction, covering (I) aneurysm location and characteristics, (II) type of procedures performed with access site, (III) procedure time, (IV) complications, (V) robot device used in the investigation, (VI) fluoroscopy time, (VII) radiation time, and (VIII) origin and date of publication of the study, was also independently performed by (ANONYMIZED FOR REVIEW), with conflicts resolved by the third author (ANONYMIZED FOR REVIEW). We conducted a manual search of references from included papers and review articles to identify potentially eligible studies that might have been missed by our database searches.

Results

The search was conducted on October 31, 2023, resulting in the identification of 109 articles from PubMed, 499 from Scopus, 102 from Embase, 109 from Web of Science, 7 from Cochrane Library, 36 from EbscoHost, and 133 from Ovid. In total, 995 articles were initially retrieved. Upon importing into ZOTERO, the software flagged 263 duplicates, which were subsequently removed before the screening process. Independently, (ANONYMIZED FOR REVIEW) screened 732 records, leading to the identification of 16 works that were deemed potentially eligible for the study. Our abstract screening process revealed 716 studies not directly relevant to in-human cerebral aneurysm intervention. These excluded works included studies focusing on ex-vivo and phantom models, review articles, and any studies that did not meet our predefined inclusion criteria.

All works eligible for review were retrieved. Ultimately, 6 works from the database search were included, while others detailing angiography procedures (8), microscope assist (1), and an additional abstract (1) were excluded. Through a manual reference search, we identified 1 work that met our inclusion criteria. Finally, we have included 7 works in this systematic review [8,9,10,11,12,13,14]. The full search strategy is depicted in Fig. 1 [7].

Fig. 1
figure 1

PRISMA flow diagram, made with PRISMA template [7]

Full size image

We have identified six original works focusing on the application of robotic devices for surgical interventions in aneurysm cases. Additionally, one work was identified as a follow-up to another study included in this review; we have included this work as part of the original investigation (Table 1) [8]. Four of these works originated from a US-Canadian collaboration, one was from the USA, one from France, and one from international collaboration. The works span various publication years, including three from 2023, and one each from 2022, 2021, 2020, and 2007. The robotic systems utilized in these works varied, with four employing the CorPathGRX Robotic System, one used StereoTaxis Telstar Magnetic Navigation System, and one used WalterLorenz Surgical Assist Arm. For detailed information regarding study design and patient characteristics, refer to Tables 1 and 2, respectively.

Table 1 Characteristics of included works
Full size table
Table 2 Patient demographics from the studies
Full size table

CorPath GRX system

CorPath GRX System (Corindus, Siemens Healthineers, USA) is a next-generation device, replacing the former version CorPath 200. The system is designed for vascular interventions (percutaneous coronary, peripheral and neuroendovascular) [8, 9, 12]. CorPath consists of a bedside unit and a remote steering unit for the physician (Fig. 2 [15]). The bedside unit is made of a robotic arm with drive and a sterile, single-use cassette for the neurosurgical devices. Arterial access and the placement of a guide catheter are done in a conventional way. Devices that will be used in the neuroendovascular interventions (i.e., coiling, flow diverters, intrasaccular devices) are loaded into the appropriate tracks of the cassette. The main components of a robotic cassette are the sheath retainer and support track at the tip of the device and torque control for guide catheter through a dedicated path. CorPath GRX is compatible with 0.0014- or 0.0018-inch guide wires, rapid exchange catheters and more. The device is controlled through joysticks behind a radiation-shielded workstation by a qualified specialist thanks to the mechanical transmission module. This allows for highly precise movements of the device. Real-time visualisation allows for monitoring fluoroscopy and haemodynamics parameters during surgery. Devices can be inserted, retracted, rotated and moved during the operation. Rotation is available with catheters and wires in increments of 1 mm. Additionally, the device possesses an automated contrast substance injector. Active Device Fixation software allows for increased precision of the devices, which is especially beneficial during neuroendovascular procedures. The device can also create cervical and intracranial angiograms when the wireless fluoroscopy pedal and power injection to the diagnostic catheter are connected.

Fig. 2
figure 2

Reprinted with permission of Siemens Healthineers [15]

CorPath GRX system in action - University Hospital Giessen performs first robotic-assisted coronary intervention in Germany.

Full size image

The first robot-assisted cerebral aneurysm operation utilizing this robotic device was documented by Pereira et al. [11]. The patient, in the 60s with a 12 mm x 11 mm lesion, underwent the procedure due to the challenging anatomical configuration of vessels. The precarious location of the aneurysm, near the tip of the basilar artery, posed a hindrance to stent placement. Prior to the operation, the attending physician and the surgical team invested over 30 h in acquainting themselves with the system. The patient, under general anaesthesia supplemented with heparin, underwent the manual phase of the procedure, involving the insertion of a 6 F, 0.088-inch inner diameter, 90 cm femoral sheath (Neuron MAX 088) into the right subclavian artery.

Subsequently, an intermediate catheter (Sofia 6 F Microvention) was navigated to the right vertebral artery (V4 segment). Simultaneously, the CorPath GRX was prepared, with a 1.7 F microcatheter (Excelsior SL-10) and a 0.014-inch microwire (Synchro Stryker Neurovascular) loaded into the cassette. The robotic arm was then employed to advance the microwire to the right posterior cerebral artery (P1), facilitating the introduction of a 4.5 mm x 21 mm Neuroform Atlas nitinol self-expanding microstent (Stryker Neurovascular) into the distal basilar artery via the microcatheter. Retraction of the microcatheter to the proximal vessel allowed for the deployment of the stent. Following this, a guidewire was once again utilized to navigate to the aneurysmal sac, enabling the introduction of the catheter. The exchange of the guidewire for embolic coils facilitated the filling of the aneurysmal lumen. Post-procedure angiography confirmed the success of the operation, and the devices were safely retracted from the vessels. The entire procedure lasted 2 h and 9 min without any complications, leading to the patient’s same-day discharge. A subsequent MR angiogram, conducted two weeks later, revealed the complete obliteration of the aneurysm.

Following the successful initial case, subsequent procedures within the same institution, comprising 6 cases [8, 9], were also free from complications. Preoperative preparations demanded over 100 h of familiarisation with the system, including training on patient-specific 3D printed models. The primary neurointerventionist, Pereira, operated from a radiation-shielded cockpit while utilizing the mobile robotic arm. Each procedure, performed under general anaesthesia with systemic heparin administration, involved the precise control and deployment of microcatheters, wires, stents, and coils by the robotic device. Post-procedure, patients were maintained normotensive and monitored for 24 h.

The aneurysms were localised in the basilar artery (3 cases), PComA (2 cases), and ICA (paraophthalmic segment) in one case. Flow diverters were deployed twice, while the stenting procedure was performed four times. The femoral artery was chosen for access in 5 cases, while the radial artery was used once. The total average operation time was 117.3 min ± 47.3 min, with 85 min ± 15 min using the robotic system. No complications related to the robot’s usage were reported.

In their recent work, Pereira et al. outlined 117 multicentre cases from 6 countries, employing the CorPath GRX system in each instance [12]. Inclusion criteria required patients to be 18 years and older, with an unruptured aneurysm indicating the need for surgery based on specific features such as a dome-to-neck ratio greater than 1.5 or a neck width greater than 4.0 mm. The procedures were designed to be conducted both manually and with robotic assistance. The primary effectiveness endpoint was achieved if the robotic operation proceeded without any deviation from manual methods, avoiding issues in microcatheter or guidewire navigation, stent deployment, or coil embolization.

The study encompassed 117 cases, who were eligible for the procedure. Four patients did not achieve the primary effectiveness point due to major strokes occurring twice and the rupture of an aneurysm during operation in two patients. Seven procedures necessitated a switch to conventional techniques, involving two device malfunctions (one related to the inability of joystick activation and one caused by an unidentified, root-caused malfunction of the cassette) and five limitations in the procedures (three caused by the limited length of the microcatheter and one each due to the usage of off-labelled microcatheter and the inability to deploy a coil). In the postoperative 24-hour period, several adverse events occurred, including minor strokes 10 times, major strokes and aneurysm ruptures 2 times each (related to the primary safety point), vascular access site complications 4 times, and electrolyte disorders 2 times. Additionally, hydrocephalus, vasospasm, respiratory failure, and urinary infection each occurred once. The mean duration of the total procedure and the robotic part of the procedure was 117.3 min ± 47.3 min and 59.4 min ± 32.6 min, respectively. The operational success reported in Pereira’s study was lower than in conducted randomized controlled trials such as HELPS, MAPS, or Cerecyte [16,17,18]. In these trials, the success rate ranged between 96.6% and 97.2%, while Pereira achieved a success rate of 94% (110 out of 117 patients).

Chivot et al. (2023) detailed their experience as the inaugural European study employing this system for the treatment of cerebral aneurysms [10]. This retrospective study, conducted at a single centre, involved 10 patients who underwent embolization procedures with flow diverter stent placement between April and October 2022. Among the cases, nine were associated with cerebral aneurysms, while one pertained to a cervical aneurysm. Paraclinoid aneurysms were prevalent in six cases, with two involving AComA aneurysms and one featuring a PComA aneurysm.

Before the surgical procedures, surgeons received specialised 3-day training. Notably, one patient with an AComA aneurysm underwent treatment using two access points—employing femoral artery access for deploying the flow diverter and radial artery access for coil deployment. In three instances, a flow diverter alone was deployed, five cases involved a combination of a flow diverter with coils, and one case utilized a flow diverter with the Woven EndoBridge. Although none of the flow diverters encountered failures, complications arose during guidewire navigation in the PComA case due to the torquability limitations of the device. Consequently, surgeons opted for manual wire guidance, while the catheter and flow diverter were positioned with robotic assistance. In the case of a large ICA aneurysm with intricate anatomy and rising resistance, the robot autonomously paused, but after device rebooting, the procedure resumed without further issues. The entire operation averaged 119 min, with the robot-specific procedure taking an average of 85 min. No complications related to the device were observed, although a patient with a PComA aneurysm experienced an ischemic infarction in the affected area due to compression of the segment caused by the mass effect.

Stereotaxis Telstar magnetic navigation system

The Stereotaxis Telstar system comprises a biplane fluoroscopy device, a magnetic field generator, and a magnetic-tipped guidewire (Fig. 3 [19]). The magnetic field is controlled through computer software, offering the added benefit of visualization via a computer system. Once the guidewire is manually introduced into the desired vessel, the surgeon gains precise control over the orientation of the guidewire’s tip through a computer interface.

The distal tip of the guidewire undergoes directional changes facilitated by an increase in homogeneous magnetic fields. This capability is attributed to the presence of a magnet housed in a rounded cup at the distal tip of the microguidewire. The rounded configuration of the cup serves to minimize potential complications during the neuronavigation process. Notably, the magnet itself, measuring 0.079 inches (ca. 2 mm) and composed of neodymium, iron, and boron, enables the distal tip to achieve a deflection angle of 130° at any angle, with a minimum radius of 6 mm during flexure. The maximum magnetic power field output is 0.15 T. Enhancing radiopacity, providing strain relief, and ensuring extra durability for the microguidewire is a platinum coil integrated into the distal tip. The overall length of this wire is 210 cm. Both the introduction and retraction of the device are manual processes, facilitated by a hydrophilic coating that streamlines these maneuvers, ensuring procedural ease.

Fig. 3
figure 3

Telstar Magnetic Navigation System, reprinted with Open Access license [19]

Full size image

Dabus et al. (2007) conducted a study, marking the first use of this magnetic-guided device in aneurysm surgery [13]. Over a period spanning April 2001 to March 2003, a cohort of ten consecutive patients was selected for this investigation. Predominantly, nine patients grappled with diverse cerebral aneurysms: two with AComA, two with basilar tip, two with hypophyseal, and one each with basilar, cavernous, and ICA bifurcation aneurysms. Additionally, a patient afflicted with mandibular arteriovenous malformation was included in the study, undergoing an embolization procedure.

In nine instances involving cerebral aneurysms, varied procedures were administered: six underwent aneurysm coiling, while in three cases, artery sacrifice procedures were performed. Notably, in the case of a 12 mm ICA bifurcation aneurysm, attempts at coiling proved unsuccessful despite a successful microcatheterization procedure.

In the majority of cases, a 5 F/6F Envoy guide catheter was employed, with the exception being the giant ICA cavernous artery, where a 7 F Meditech balloon catheter took its place. The microcatheters utilized in these cases were Prowler Plus and Excel-14. Throughout the procedures, the microguidewire system was navigated via a touchscreen tablet. The magnet in the guidewire took approximately 10 to 20 s to power up during the procedure. The device adeptly advanced to the desired target, and upon reaching its final point, a microcatheter seamlessly followed suit. Subsequent angiography procedures assessed for any potential lesions. The ensuing stages of the surgery adhered to conventional procedures.

All procedures across all cases were conducted via the femoral artery, achieving the desired vessels successfully on the first attempt. Remarkably, there were no reports of serious neurological complications or mortality. The researchers observed an absence of complications with other devices attributed to the electromagnetic field.

Zimmer biomet WalterLorenz surgical assist arm

Lastly, the Zimmer Biomet WalterLorenz Surgical Assist Arm is considered in this systematic review. This versatile arm is specifically designed for surgical retraction and instrument positioning during various surgical procedures. The device comprises an end effector, distal arm, proximal arm, and a control box with an interface system. Additional modules can be mounted to accommodate different types of procedures.

Yeung et al. applied this device during aneurysm surgical clip** procedures to assess its efficacy in managing cerebral intraoperative aneurysm rupture [14]. In a case involving the pericallosal artery, a rupture occurred during the surgery. The device, loaded with a temporary clip before the operation, was set up in less than 3 min. When bleeding started, the robot promptly loaded the clip, achieving immediate haemostasis. The aneurysm was successfully managed with three Mizuho aneurysm clips, and no further complications were noted.

Discussion

The treatment of cerebral aneurysms can vary depending on factors such as size, location, and the patient’s overall health. There are two main methods of treating cerebral aneurysms: surgical clip** and endovascular coiling. Both procedures aim to prevent the rupture of the aneurysm, which can be life-threatening.

Robots prove their usefulness in interventional medical specialities by adding competencies to physicians, both in terms of decision-making accuracy and planning, as well as the ability to reach the target in a less invasive manner and perform surgery more precisely. If imaging of the intervention site requires potentially harmful radiation exposure for medical personnel, the advantage of a telemanipulator is the ability to keep the team away from the operating table, reducing the time spent in harmful conditions. In the case of procedures that fall under life-saving actions, the requirements for such a system to meet are growing rapidly. Neurointervention seems to be an extremely challenging and crucial task, from diagnostics and planning to the actual intervention.

Currently, robots are being attempted to be used for procedures that have been successfully invented and performed without them. There is certainly value in approaching these challenges with completely new technical and medical solutions. The review shows that there is currently a lack of groundbreaking inventions, and the robotic procedures applied so far utilize the advantages mentioned earlier but do not have a breakthrough impact on the problems faced by the population at risk of death due to misdiagnosis or the inadequate removal of the danger in a timely and efficient manner. However, noticeable progress and increasing interest in using robots in this field in the near future will likely change this landscape, and entirely new surgical techniques and tools can be expected.

The introduction of automation for certain robot tasks, as well as the increased efficiency of operating a telemanipulator at various distances from the patient, are associated with the need to implement sensors that allow for measurable assessment of the tool’s performance in the vascular field. This data can be utilized for optimization and enhanced safety during surgery.

Progress in automation and increased safety can be achieved if artificial intelligence elements are introduced both in the diagnostic and planning process, as well as in monitoring the work of intervention tools - this cannot be done without efficient sensing.

Current robotical systems are not without limitations. Pereira and Chivot reported that present devices can manage only one catheter and microwire at a time [10, 12]. At the moment, performing the jailing technique with robotic devices is not possible, as two catheters cannot be manipulated [20,21,22]. This creates a problem with aneurysms larger than 10 mm, as it significantly increases procedure time - one catheter has to be guided after the previous one is done [10]. Notably, Chivot reports a lack of haptic feedback in the robotic system. While an experienced operator can compensate for the lack of this feature through visualization on screens, a less-experienced practitioner may encounter issues initially.

To the best of the authors’ knowledge, this is the very first study dedicated solely to the topic of intracranial aneurysms. While Crinnion et al. previously published a systematic review on the application of neurosurgical robots for neurointerventions [23], their study provided a general overview of robotic applications, without focusing on any specific pathology. Their review included one work related to the aneurysm - Pereira’s initial study [11]. In contrast, our study offers an updated and in-depth exploration of the current state of robotic devices exclusively within the context of cerebral aneurysms.

This work is not without limitations. Our review process employed web-based translators for studies not originally published in English or Polish. Secondly, our focus on clinically applicable interventions may have limited our ability to capture the latest technological advancements within the field. While this review aimed to assess current clinical applications, acknowledging promising technologies identified in pre-clinical studies using phantom/animal models could provide valuable insights for future research directions.

Recent technological advancements include the use of small-scale robotic devices incorporating shape memory alloys and biocompatible polymers for intervention, navigation with magnetic particle imaging for improved visualization, and the application of artificial intelligence methods for aneurysm rupture prediction [24,25,26]. Finally, testing of remote aneurysm neurointerventions, may provide a background for long-distance surgeries in the near future [27]. Exploring the clinical potential of these emerging technologies in future studies could be particularly useful.

Conclusions

In summary, this systematic review focuses on the application of robotic devices in cerebral aneurysm neurointerventions. While showcasing the advantages of reduced radiation exposure and enhanced precision, we emphasize existing challenges, including limited number of studies, variable success rates, and technical constraints. Ongoing advancements, particularly in jailing procedures and haptic feedback, as well as the application of artificial intelligence, are vital for the future integration of robotic devices in intracranial aneurysm surgeries.