A Comparative Study for Control of Semi-Automatic Robotic-Assisted Ultrasound System in Spine Surgery

Abstract

Ultrasound (US) imaging has been widely applied in different clinical scenarios thanks to its low-cost and non-radiative nature. Recently, robotic US has increasingly been investigated to produce 3D US reconstructions for navigation during surgical interventions. Robotics is considered of interest to address variable skills among human sonographers. Dedicated control strategies are needed to ensure high-quality robotic US reconstructions that are comparable or superior to those generated by human experts. The robot controller ought to establish human-like scanning maneuvers while maintaining tight skin contact and ensuring essential safety. In essence, this means that at all times, the robot should ensure contact while avoiding the application of excessive force on the patient. To acquire an improved understanding of what is an optimal control method, a comparative study on several admittance-based controllers was conducted while a semi-automatic path planning approach was used to realize automatic US scanning. The developed system was validated by scanning a synthetic phantom, compared with position and admittance control. The robotic US system with the proposed control applied a force lower than <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$3.83\pm 0.31\ \mathrm{N}$</tex> while ensuring continuous US imaging. By defining the successful rate as a US image has more than 90% of soft tissue (i.e., ligament) length, the velocity-based admittance controller has more than 80% successful US imaging. Such an approach could contribute to the further development and uptake of robotic US systems in spine surgery or applications.