Autonomous Path Planning for Vascular Interventional Surgical Robots: From Virtual Simulation to Physical Validation

An autonomous navigation framework for vascular interventional surgical robots is proposed in this study, with a focus on path planning based on virtual vascular models and real-time execution control. A complete navigation pipeline is developed, including 3D vessel reconstruction, centerline extraction, fully connected graph construction, and A * search-based optimal path generation. A set of geometric quality metrics, such as path length, curvature, and vessel compliance are used to quantitatively evaluate the feasibility of each planned path. The navigation system is implemented in the SOFA simulation environment with real-time feedback control, and further validated on a physical robotic platform. Experimental results from six navigation trials under different target settings demonstrate that the proposed method enables stable, accurate guidewire advancement with high success rate. The system effectively bridges virtual planning and physical control, providing a robust foundation for intelligent robotic navigation in vascular intervention.