Search

Abstract: Robot-assisted arthroscopic surgery has been receiving growing attention in the field of orthopedic surgery. Most of the existing robot-assisted surgical systems in orthopedics take more focus on open surgery than minimally invasive surgery (MIS). In traditional arthroscopic surgery, a specific type of MIS, the surgeon needs to hold an arthroscope with one hand while performing the surgical operations with the other hand, which can restrict the dexterity of the surgical performance and increase the cognitive load. Additionally, the surgeon heavily relies on the arthroscope view when conducting the surgery whereas the arthroscope view is a largely localized view and lacks depth information. This motivates us to develop a two-arm robotic system, a robot-assisted arthroscope holder, and a robot-assisted surgical tool with haptic feedback, to assist the surgeon in both scenarios. In a robot-assisted system for arthroscopic surgery, surgical tools attached to the robot end-effector (EE) will affect the robot dynamics inevitably, which could undermine the utility and stability of the robotic system if the dynamic uncertainties (e.g., the mass of the surgical tools) are not identified and compensated for in the robot control system. To solve this problem, an integrated framework of impedance control and nonlinear disturbance observer (NDOB) is proposed, where the former ensures compliant robot behavior and the latter compensates for dynamic uncertainties. By integrating an impedance controller with NDOB, the proposed framework allows an accurate impedance control and stable system when dynamic model inaccuracy and external disturbance exist. However, the NDOB always estimates all of the uncertainties as a lumped term, and it is not able to separate any of the components. As a further step, we developed a framework for using a neural network (NN) to learn for some uncertainties thus separating the other uncertainties. The effectiveness and performance of the trained NN model are verified in simulations. However, it is not an efficient approach in practice considering the laborious offline training procedures. Aiming for a more compact and efficient approach, we developed a gravity iterative learning (Git) scheme with a steady-state scaling strategy specially for gravity compensation. The Git scheme can accurately learn and compensate for gravity when gravity compensation is the main concern. By integrating the Git scheme with an impedance controller, the robot can keep still at a designated pose even with a heavy payload attached to the robot EE. Also, it allows the operator to move it freely via a pedal switch whenever needed. On the other hand, virtual fixture (VF) has been serving as a vital role in robot-assisted surgeries, such as protecting a beating heart. In orthopedic surgery, preoperative images are often used in the operating room, on which some curves can be drawn to mark out the boundaries for osteophytes to be removed. These curves can be used to generate VF to assist in removing osteophytes during the operation. A challenge is that the hand-drawn curves usually have irregular shapes and cannot be mathematically represented by equations, thus most of the existing VF approaches will not work in this scenario. To this end, a point-based VF-generating algorithm is developed, with which 3D VF can be generated directly from point clouds in any shape including the hand-drawn curves in a preplanned image. In the end, a prototype of a two-arm robot-assisted system for arthroscopic surgery is built and preliminarily evaluated. The left-arm robot is used as a robot-assisted arthroscope holder, which can hold the arthroscope still at any designated pose, while also allowing the operator to move it around freely via a pedal switch whenever needed. The left-arm robot is implemented with an impedance controller and the Git scheme, where the former can provide compliant robot behavior thus ensuring a safe human-robot interaction, while the latter can accurately learn and compensate for gravity. The right-arm robot is used as a robot-assisted surgical tool providing VF assistance and haptic feedback during the surgery, and is implemented with the point-based VF algorithm, which can generate VF directly from point clouds in any shape, render haptic force feedback, and deliver it to the operator. Furthermore, the VF, the bone, and the surgical tool with its real-time position are visualized in a 3D virtual environment as additional visual feedback for the operator.