Résumé
This work presents a closed-loop strategy for 3D online motion planning of beveled steerable needles using duty-cycled rotation. The algorithm first selects an entry point that minimizes a multi-criteria cost function and then combines an RRT-based path planner with an intraoperative replanning algorithm andworkspace feedback information to constantly update the needle inputs and adjust the trajectory. Simulations in a workspace based on a typical prostate needle steering scenario showthat the algorithm is robust against disturbances and model uncertainties and can provide online trajectories to avoid obstacles even under the presence of physiological motion.