The Future of Robot-assisted Spine Surgery
The culture of constant innovation and technological advancement is a factor that attracts many to pursue a career in neurosurgery. On the forefront is robotic assistance in the operating room.
Spine Robotics Today
The current generation of neurosurgical robots is focused on trajectory assistance for navigated implants.1
The ROSA (Zimmer Biomet, Warsaw, Ind.) robot is typically used in epilepsy surgery and helps the neurosurgeon place multiple stereo-EEG leads or other navigated cranial implants.2 Multiple trajectories are planned on the preoperative MRI. After registering the navigation system, the robotic arm moves an instrument guide aligned with the correct trajectory to the entry site. The surgeon then places the hardware through the guide. This system has also been adapted for screw placement.3
The Mazor Rennaisance (Mazor Robotics, Israel) is the 2011 successor to the 2004 SpineAssist robotic platform used for navigated screw placement. This system involves a track that is surgically fixed to the spine. Intraoperative fluoroscopy is used to register the robot’s navigation software to the preoperative CT scan. A robotic guidance unit, which contains an instrument guide, is then attached to the track. The guidance unit then aligns to the correct trajectory, allowing the neurosurgeon to drill and place screws through the instrument guide. The 2016 Mazor X is the most recent version of this robot series; it transitions to a robotic arm-mounted instrument guide, uses intraoperative fluoroscopy-to-CT registration and incorporates 3-D surface scanning.
The most recent advancement in robot-assisted screw placement is the ExcelsiusGPS (Globus Medical, Audubon, Pa.), which was FDA-approved in 2017. This system provides multiple options for registration, including registration to a preoperative CT with intraoperative fluoroscopy, intraoperative CT or intraoperative fluoroscopy.4 Screw trajectories are planned on the robot’s software interface. A robotic arm mounted to a floor unit places a guide to pre-planned trajectories, while the neurosurgeon performs drilling and screw placement through the aligned robotic instrument guide. The instrument guide is removable, allowing for potential additional tools to be mounted to the robotic arm.
Trajectory assistance was a natural target for the first generation of spine robots. The workflow for these robotic systems will continue to be refined as they become more widely used in clinical practice.
Spine Robotics in the Near Future: Application of Existing Technology
An unmet area of growth is the use of robots to assist with bone removal for decompression (e.g. laminectomies and facetectomies) and deformity cuts (e.g. posterior column and pedicle subtraction osteotomies). Although this is a futuristic application, the technology already exists.
Most of our technological needs for a robot capable of performing laminectomies and osteotomies can be found in the arena of computer-aided manufacturing (CAM), which has been used in the automotive and aircraft manufacturing industry for decades. Asking a robot to remove a specific area of bone is akin to an industrial robot performing a milling operation for subtraction manufacturing.
Let us consider the specifics of the standard use of robotics for milling in the manufacturing industry. For example, in the production of a gear from a steel block, a seven-axis robotic arm moves through a complex pattern without collision to transform the steel block into a gear. This process, requiring computer instructions that are optimized to the specific starting and ending geometry, is known as computer numerical control (CNC).
This optimization problem is readily solved by existing CAM software packages, such as PowerMILL, FeatureCAM and Robotmaster.5-7 The starting and ending geometries and robotic arm details are imported into the software, then the pattern of robotic arm movement required to produce the given product is generated.
In the imagined spinal operating room of the near future, the neurosurgeon will define the 3-D regions of bone removal on a preoperative or intraoperative CT scan. A CAM software package will automatically determine the pattern of robotic arm movement necessary to remove a bone. Present day neurosurgeons have utilized this principle by robotically performing a trans-labyrinthine approach in a cadaver in two minutes and 30 seconds.8
A Promising Application of Artificial Intelligence
The forefront of robotics research is its integration with an artificial intelligence technique called machine learning. In machine learning, the robot’s actions are governed by concepts it learns by example, similar to the way humans learn. As an example, the Autolab group at UC Berkley trained a surgical robot to cut complex patterns out of deformable gauze.9 A grasping arm has to constantly adapt the grasp position, tension direction and magnitude, while the cutting arm changes directions as it cuts the pattern. If done incorrectly, the gauze deforms. A camera detects gauze deformation and automatically improves the behavior of the tensioning arm. Accordingly, spine robots of the future will likely integrate artificial intelligence algorithms to automatically adapt to changes in the operative environment. As the thecal sac is decompressed, the dura will move to a potentially unsafe location. The robot will need to recognize this by computer vision and dynamically adapt its trajectory to avoid the new danger zone.
The current generation of robots in spine surgery is focused on automatically aligning trajectories for screw placement. In the near future, robots will assist the neurosurgeon in performing laminectomies and osteotomies by adapting computer-aided manufacturing technology already utilized in the automotive and aeronautics industries. Subsequent developments will integrate artificial intelligence algorithms to assist robots to anticipate the dynamic changes during surgery. Robotic platforms have the potential to increase accuracy and efficiency during mechanical tasks, but will not completely replace the unparalleled ability of human judgement used to adapt in unexpected situations during surgery. The manufacturing industry has demonstrated that optimal accuracy and efficiency is achieved when robots and humans work side-by-side, each performing tasks they do best.
1. Theodore, N., Arnold, P. M., & Mehta, A. I. (2018). Introduction: The rise of the robots in spinal surgery. Neurosurgical Focus.
2. De Benedictis, A., Trezza, A., Carai, A., Genovese, E., Procaccini, E., Messina, R., . . . Marras, C. E. (2017). Robot-assisted procedures in pediatric neurosurgery. Neurosurgical Focus.
3. Joseph, J. R., Smith, B. W., Liu, X., & Park, P. (2017). Current applications of robotics in spine surgery: A systematic review of the literature. Neurosurgical Focus.
4. Zygourakis, C. C., Ahmed, A. K., Kalb, S., Zhu, A. M., Bydon, A., Crawford, N. R., & Theodore, N. (2018). Technique: Open lumbar decompression and fusion with the Excelsius GPS robot. Neurosurgical Focus.
5. PowerMILL (2018). Expert high-speed and 5-axis machining software. Autodesk, Inc. Retrieved from https://www.autodesk.com/products/powermill/overview
6. FeatureCAM. (2018). Make parts faster with automated CAM. Autodesk, Inc. Retrieved from https://www.autodesk.com/products/featurecam/overview
7. Robotmaster. (2018). Robotmaster CAD/CAM robotic software. Hypertherm, Inc. Retrieved from https://www.hypertherm.com/en-US/hypertherm/robotmaster/robotmaster-cadcam-robotic-software/?region=NART
8. Couldwell, W. T., Macdonald, J. D., Thomas, C. L., Hansen, B. C., Lapalikar, A., Thakkar, B., & Balaji, A. K. (2017). Computer-aided design/computer-aided manufacturing skull base drill. Neurosurgical Focus.
9. Thananjeyan, B., Garg, A., Krishnan, S., Chen, C., Miller, L., & Goldberg, K. (2017). Multilateral surgical pattern cutting in 2D orthotropic gauze with deep reinforcement learning policies for tensioning. 2017 IEEE International Conference on Robotics and Automation (ICRA).
No upcoming events
Be the first to reply using the above form.