Robotic neurointervention: Beyond the noise

Despite the justified excitement surrounding its potential, the integration of Robotic Neurointervention (R-NI) into clinical practice remains unrealized. Benchtop demonstrations focusing on isolated procedural components fail to address the practical challenges hindering real-world R-NI implementation. These demonstrations, often widely shared on social media, generate considerable hype but can be misleading. A critical examination of the barriers to R-NI implementation and the limitations of current R-NI technology is essential to advance the field.

The most extensive R-NI study to date, involving 130 patients with unruptured intracranial aneurysms, provides valuable insights into the current state of the technology. After excluding 9 screen failures and 4 device malfunctions, 117 patients underwent R-NI procedures. Notably, all cases required manual placement of the access system by a Neurointerventional clinician before transitioning to robotic control for microcatheter and microwire or device manipulation. While 110 patients (94%) completed their robotic-assisted procedures¹, this study highlights several barriers to wider adoption. These challenges include the requirement for a neurointerventional clinician to manually position the access system and the need to convert back to manual procedures due to the technological limitations of R-NI.

An important barrier to adoption is that robotic neurointervention (R-NI) systems have yet to demonstrate benefits for patients. While technical outcomes equivalent to historic literature controls have been shown for completed R-NI procedures, ¹ this comes at the cost of additional human resources, reduced angiography lab efficiency, and unclear patient benefit. R-NI may reduce radiation exposure for physicians, but this benefit might be offset by extended procedure durations, inadvertently increasing radiation exposure for patients and other angiography lab staff. ²

What does R-NI need to achieve widespread on-premises adoption?

While R-NI's ability to expand access to time-critical interventions through telerobotic NI (TR-NI) is widely acknowledged, its potential impact during on-site implementation should not be overlooked.

A primary goal for R-NI should be reducing vascular complications caused by endovascular tool tip contact with the vessel wall. Current manual control methods often lead to inadvertent tool tip contact with the vessel wall, potentially resulting in injury and thromboembolic complications. Robotics could address this issue by optimizing the multidirectional control of tools, thereby preventing traumatic vessel wall contact and reducing vascular injury. This could also potentially reduce rates of asymptomatic microemboli and symptomatic post-procedure ischemic stroke.

Robotic technology has the potential to simplify complex endovascular navigation. Enhanced multidirectional control over endovascular tools with improved visualization of imaging could reduce overall procedural times, increase procedure completion rates, and minimize radiation exposure for all angiography lab staff—not just the neurointerventional clinician.

To gain broader acceptance, R-NI systems need to demonstrate efficiency comparable to or better than manual procedures. This should be assessed holistically, from case initiation to the start of the next, with a particular focus on robotic setup and takedown times. In clinical practice, if R-NI is so time-intensive that it reduces the number of cases performed from three to two in a morning list, adoption will be limited. Additionally, systems requiring the constant presence of a neurointerventional or other skilled clinicians at the bedside, in addition to the R-NI operator, will face challenges integrating into routine clinical practice.

Finally, R-NI offers an opportunity to advance Neurointervention beyond the current standard of care and set new benchmarks for patient safety. Incorporating artificial intelligence into R-NI systems could enhance intraoperative decision-making, leading to more effective interventions and improved patient outcomes.

What R-NI should look like

For R-NI to become a tool routinely used to enhance patient care, it should be capable of performing entire procedures. Systems lacking this capability are likely to disrupt current clinical workflows and become underutilized. R-NI must be capable of navigating from the access vessel to the pathology site and managing multiple tools simultaneously in the process. Users should be able to adjust settings and inject radiopaque contrast. Additionally, R-NI systems must reliably navigate, deploy, and retrieve end-effector tools. Precise control of microcatheters and guidewires is crucial; a robot failing to instill confidence in this area may be dangerous and unlikely to gain widespread adoption.

What TR-NI should look like

Telerobotic-assisted endovascular maneuvers, such as bi-axial guidewire manipulation and stent deployment, were first published in 2019. ³ However, a complete telerobotic TR-NI procedure has yet to be performed and published in the peer-reviewed literature. While latency and connectivity issues are often cited as the main barriers to TR-NI, remote control of an endovascular robot without a 5G network has been demonstrated with acceptable latency. The primary obstacle to TR-NI is that existing R-NI technology cannot yet perform a complete endovascular procedure. Beyond this core capability, additional criteria must be considered to successfully deploy and achieve widespread adoption of TR-NI.

Operational equivalence

The transition to TR-NI will require remote operators to possess robotic capabilities at least equivalent to or better than manual NI. Ensuring operational equivalence between manual and TR-NI procedures mandates that R-NI systems and operators can effectively manage intra-procedural complications remotely, such as an intracranial vessel rupture during mechanical thrombectomy.

Clear approach to management of latency and connectivity

While not the primary obstacle to TR-NI, managing latency and connectivity is important. There should be well-defined and justified thresholds for unsafe latency, reliable systems for intraoperative connection monitoring, robust processes for extended preoperative connection testing, and effective methods for handling unsafe latency and connection loss.

Visibility, audiovisual communication, and imaging integration

High-quality audiovisual communication between the remote neurointerventional clinician and the onsite R-NI support team is crucial. Remote users should have the same or enhanced procedural visibility compared to manual or on-premises neurointervention. This includes a clear view of all live imaging feeds, hemodynamic monitoring, and high-quality communication with the bedside support staff and the robot. Interfaces with on-site imaging equipment should provide a high-resolution display of all fluoroscopy as well as direct communication with technicians/operators controlling the C-arm and other imaging functions on-site.

Normalizing remote procedures and simplifying execution

Surgical robotic systems often require clinical staff to learn specialized skills for setup, bedside management, and tool loading and operation. This requirement significantly hinders scaling up TR-NI use in facilities with limited neurointerventional experience. Staff with minimal training in remote sites should be able to quickly and safely set up and interact with the TR-NI system for emergency procedures. For TR-NI to gain adoption, it should be as simple as possible every time it occurs.

Conclusion

Robotic systems must be designed and rigorously tested to ensure they enhance patient outcomes and augment neurointerventional clinician capabilities. We must cut through the noise and focus on advancing robotic technology to fully realize its potential impact for both on-premises and remote neurointervention. By addressing real-world challenges, R-NI can set new benchmarks in safety, expand access to time-critical care, and evolve into an indispensable tool for neurointerventional clinicians.