Autonomous Guidewire Navigation for Robot-assisted Endovascular Interventions: A Knowledge-Driven Visual Guidance Approach
Wentao Liu, Weijin Xu, Xiaochuan Li, Bowen Liang, Ziyang He, Mengke Zhu, Jingzhou Song, Huihua Yang, Qingsheng Lu
TL;DR
The paper addresses the challenge of autonomous guidewire navigation in robot-assisted endovascular interventions, where reliance on expert demonstrations limits scalability. It introduces a knowledge-driven visual guidance (KVD) framework that fuses imaging-derived vascular maps, guidewire tip localization, and a boundary-aware path planner (BDA-star) with a reinforcement learning environment and a pre-trained CNN for robust feature extraction. Key contributions include the BDA-star algorithm, an RL environment with explicit observations and a tailored path-navigation reward, and demonstrations of 100% success in simulating navigation to LSA, LCA, and BCA with improved efficiency on a real robotic platform. The findings suggest significant potential to reduce procedure times and radiation exposure, while also highlighting challenges in generalization to unknown targets and the need for realistic virtual environments for clinical translation.
Abstract
Autonomous robots for endovascular interventions hold significant potential to enhance procedural safety and reliability by navigating guidewires with precision, minimizing human error, and reducing surgical time. However, existing methods of guidewire navigation rely on manual demonstration data and have a suboptimal success rate. In this work, we propose a knowledge-driven visual guidance (KVG) method that leverages available visual information from interventional imaging to facilitate guidewire navigation. Our approach integrates image segmentation and detection techniques to extract surgical knowledge, including vascular maps and guidewire positions. We introduce BDA-star, a novel path planning algorithm with boundary distance constraints, to optimize trajectory planning for guidewire navigation. To validate the method, we developed the KVD-Reinforcement Learning environment, where observations consist of real-time guidewire feeding images highlighting the guidewire tip position and the planned path. We proposed a reward function based on the distances from both the guidewire tip to the planned path and the target to evaluate the agent's actions.Additionally, to address stability issues and slow convergence rates associated with direct learning from raw pixels, we incorporated a pre-trained convolutional neural network into the policy network for feature extraction. Experiments conducted on the aortic simulation autonomous guidewire navigation platform demonstrated that the proposed method, targeting the left subclavian artery, left carotid artery and the brachiocephalic artery, achieved a 100\% guidewire navigation success rate, along with reduced movement and retraction distances and trajectories tend to the center of the vessels.