The State of Robotics in Neurosurgery

0
13183

The use of robotics in surgical practice has been a field of rapid development during the last few decades. From simple biopsies to complex reconstructive surgeries in anatomically challenging sites, robotic surgery has been adopted by multiple specialties, allowing surgeons to push the limits of their technical skills. Neurosurgery has been slower than other surgical specialties in incorporating robotics into routine practice, owing to the anatomical complexity of neurological structures and the confined space in which neurosurgeons must operate. Regardless, the field is well-suited for the implementation of robotic technology, due to its rich history of innovation and technological application in practice, the microsurgical and highly technical nature of neurosurgical procedures and a culture of adopting and embracing new technology with optimism.1 In tandem with increased attention to minimally invasive approaches to the brain and spine, there is no doubt that robotics will transform the field of neurosurgery in the very near future.

The first application of robotic surgery was in the field of neurosurgery. In 1985, Kwoh et al. used a Unimation Programmable Universal Machine for Assembly (PUMA) 200, a machine originally designed for industrial purposes, to perform CT-guided stereotactic biopsy of a brain lesion.2 Shortly thereafter, the PUMA 200 was used as a retractor in the resection of thalamic low-grade astrocytomas in children.3 Although it has been discontinued, the PUMA 200 is considered the predecessor of current surgical robots. The NeuroMate robot (Integrated Surgical Systems, Sacramento, California) was developed in the 1990s for both craniofacial and spinal stereotactic neurosurgery.4 It marked the first FDA-approved robotic device for neurosurgical procedures. Following these pioneering efforts, several other robots, such as ROSA® (Zimmer Biomet, Montpellier, France), Pathfinder (Prosurgics, High Wycombe, United Kingdom), NeuroArm (University of Calgary, Calgary, Alberta, Canada), SpineAssist (MAZOR Robotics, Caesarea, Israel), MKM (Zeiss, Oberkochen, Germany), CyberKnife (Accuray, Palo Alto, California, US) and the Steady Hand System (Johns Hopkins University, Baltimore, Maryland), have been developed and used in neurosurgery.5,6

o

Categories of Systems and Examples

There are a variety of robotic systems commercially available for neurosurgery; they differ in the amount of interaction between the surgeon and the system and the degree of control allowed to the robot. As such, three classifications exist based on this criteria:

  1. Autonomous systems
  2. Dependent systems (master-slave)
  3. Shared-control systems

Autonomous Systems

Autonomous systems are designed to reproduce pre-programmed motions along preset coordinates decided by the surgeon. They have been used robustly in stereotactic applications, especially in biopsy guidance, depth electrode placement and, more recently, pedicle screw placement in spine surgery. The initial PUMA systems were a good example of autonomous systems. The Minerva system (University of Lausanne, Lausanne, Switzerland) is a similar platform developed as a dynamic CT image guidance system in which the surgeon can adjust the trajectory in real-time and adjust for variables such as intraoperative brain shift.7 The Pathfinder has been used for insertion of needles in intracranial biopsies and guiding drills for burr hole placement by mapping possible trajectories towards a selected target based on preoperative imaging. SpineAssist and Renaissance (MAZOR) have been utilized in spinal instrumentation and minimally invasive spine surgery.

Dependent Systems (master-slave)

Robotic systems that fall under this category are those that allow the surgeon to maintain full, remote control of the robot at all times. One such example of this system is the NeuroArm, an MRI-compatible surgical robot with eight degrees of freedom (DOF), compatibility with microsurgical instruments and haptic feedback mechanisms. The neurosurgeon may work remotely from the robot in an adjacent room. The NeuroArm working station has been used in over 1,000 cases including MRI-guided biopsies, microsurgical dissection and hematoma evacuations, with favorable results.8

Shared-control Systems

Shared-control systems are a hybrid of the two system categories described above, in which the surgeon and robot share control over the desired maneuvers; for example, a surgeon may dictate the movements of the robotic arms and instruments, while the system may subtract tremor or muscle fatigue, allowing for a more precise and finer dissection of delicate neural tissues. The Steady Hand System is one such dexterity-enhancement system designed to aid microsurgery through filtering tremor and unwanted movements. However, its clinical utility has been limited to retinal microsurgery and has yet to be trialed in neurosurgical research. ROSA® is a system utilized mostly for depth electrode placement and intracranial biopsies, limited by its lack of compatibility with microsurgical instruments. The NeuRobot (Shinshu University, Matsumoto, Japan) is a similar platform designed more specifically for use in minimally invasive procedures such as endoscopic third ventriculostomy, intracranial tumor resection and portions of other intracranial procedures such as microsurgical Sylvian fissure dissection.9

Challenges and Future Directions

The future of robotic neurosurgery is very promising. Although neurosurgery is among the last surgical fields to incorporate this novel technology, its potential utility cannot be understated. However, there are significant hurdles inherent to neurosurgery that are perhaps responsible for the delay in utilization of robot-assisted surgeries; these must be addressed prior to routine implementation.

One such difficulty is neurological anatomy that makes it difficult to utilize robotic assistance. Neurosurgical procedures have a markedly lower tolerance for small errors compared to other surgical specialties, due to the permanence of any damage to neural structures; these inherent risks may have impeded the adoption of robotic assistance into the field of neurosurgery. Spatial restrictions imposed by the anatomical confines of the cranial vault or spinal column provide very little room for manipulation of robotic arms or instruments, compared to an insufflated abdominal cavity, for example. The development of more refined robotic instruments coupled with enhanced optic visualization technology may circumvent this obstacle.

A major limitation of robotic systems is the lack of satisfactory haptic feedback, a reproduction of the natural tensile forces experienced upon traction of tissues or instruments. This sensory feedback is extremely important in avoiding intraoperative insult to fragile neural structures. Although there have been attempts to replicate this “sense of touch” in newer robotic platforms, surgeons are still limited to relying mostly on visual inputs to determine the amount of force applied to tissues. In a similar sense, there is an additional loss of proprioceptive feedback during robotic surgery; this has not been adequately addressed.

The training of the next generation of robotic neurosurgeons also poses a challenge. Mastering surgical maneuvers on robotic platforms involves a significant learning curve, which is augmented by the fact that neurosurgery residents do not have exposure to such platforms routinely during their training. Virtual reality programs that enable interaction with the robotic interface have been suggested. Trainees may also practice using robotic systems on cadaver models to gain familiarity of the workstation environment and robotic surgical approaches.

Regardless of these obstacles, robotic technology may afford the future neurosurgeon the ability to increase surgical accuracy as well as effectively and efficiently execute more complex procedures. Robot-assistance may overcome human limitations of tremor and operative fatigue, while allowing for more dynamic, real-time guidance of surgical trajectories. Although our current robotic technology is limited by a variety of technical challenges, it holds great promise as a powerful tool in the future neurosurgeon’s armamentarium.

References

View All
Print Friendly, PDF & Email
o