A Heuristic-Sliding-Window-based RRT Path Planning for Endovascular Catheterization

Abstract

Catheter interventions are often used in endovascular procedures to obviate complicated open surgical interventions. One of the major challenges relates to moving the catheter toward the required location with safety and accuracy. Due to the unpredictable tissue deformation associated with device insertion and the uncertainties of intra-operative sensing, a fast and robust path planning algorithm would be advantageous. Most of current methods are pre-operative planning, ignoring time costs. This paper aims at proposing a faster and robust path planning algorithm based on heuristics information. In this paper, a novel Heuristic-Sliding-Window-based Rapidly-exploring Random Trees (HSW-RRT) path planning algorithm is proposed for endovascular catheterization. This method keeps the catheter away from vascular edges in light of safety concerns by sampling along the centerline. Simulation results show the feasibility of this path planning method in 2D scenarios. Path solutions can be generated with similar performance and less time effort than RRT*.