A Low-Cost Research Platform for Robotic Assistive Neurosurgery

Introduction: Assistive technology has been rapidly advancing in the field of neurosurgery, particularly in the past 20 years, bringing us one step closer to performing robot-based automated image guided surgery (aIGS). This work has been explored by existing robotics companies and researchers, including the ExcelsiusGPS (Globus, Philadelphia, PA) which almost places pedicle screws automatically, only relying on the surgeon for tissue manipulation and trajectory confirmation. While these technologies augment surgery in high-resource settings, low resource environments could benefit from access to tools that make pedicle screw placement possible, even in the absence of a neurosurgeon. In this work, we assess whether a low-cost robotic platform can be used to research and develop appropriate-accuracy neurosurgical procedures, and outline a method for constructing and testing such a system.

Methods: An open source AR4 robot (Annin Robotics, Nampa ID) was constructed for $2,500 for assessment of automated image guided pedicle screw placement based on prior methods. Once built, the robot was tested for repeatability and accuracy of the tool end to determine if it will be accurate enough for use in surgical settings. A test fixture using a precision optical breadboard was designed, and 3D printed markers were designed to rigidly mount to the tool end of the robot. With knowledge of the exact location of the markers relative to the base of the robot, two accuracy assessments were conducted: repeatability of returning the robot to its original position, and accuracy of movement from one position to another. These assessments were repeated 30 times each to obtain the results.

Results: Repeatability of the robot with respect to an individual test fixture marker was assessed at 0.687mm ± 0.340mm and accuracy relative to multiple was assessed at 0.705mm ± 0.422mm.

Conclusion : The system tested in this initial benchmark appeared to be accurate enough for use in surgical settings. A phantom study is currently underway and will be utilized to assess automated pedicle screw placement accuracy before an ex vivo cadaver study with this system.