Design optimization of medical robotic systems based on task performance metrics: A feasibility study for robotic guided vascular laser treatments

While designing an end-effector tool for a specific task using an already available robotic arm is a common strategy for developing robot-based solutions, it is subject to the designer's subjective biases, especially in the medical robotics field. Consequently, this approach can result in suboptimal solutions that may limit the performance of the robotic arm and the overall system. This work introduces a novel framework for developing a virtual scenario that enables the robust optimization of design guidelines for robotics-based medical solutions. The proposed framework aims to address the issue of subjectivity involved in the design process, improving the reliability and accuracy of the robotic system. Specifically, it aims to find the optimized 3D positioning of the Tool Center Point and the robot base that takes full advantage of the robotic arm kinematics. The feasibility of the proposed methodology is validated for designing a system for robotic-guided vascular laser treatments. First, a simulated treatment environment is constructed to compute the optimal trajectories to treat a wide population of synthetic human models. Next, a design optimization methodology based on robot task performance metrics is developed. Overall, the proposed design guidelines allowed the performance of 94.1% of the treatment trajectories free of collisions with median precision of around 0.5 mm while guaranteeing high manipulability. Thus, proving the added value of the proposed strategy and showing the potential for promoting the development of optimized robotics-based solutions.