Reverse-Rapidly Exploring Random Tree Based Path Planning With Insertion Pose Finding for Robot-Assisted Flexible Needles

Abstract

Abstract. In percutaneous puncture surgery, achieving precise access to the target region while avoiding anatomical obstacles remains a critical challenge. However, due to the nonholonomic constraints of the needle, planning appropriate obstacle-avoiding paths remains challenging. To address these issues, this study proposes a novel path planning method for flexible needle puncture robots operating within tissue regions. The core contribution is to select random points generated by rapidly exploring random tree (RRT) as turning points for unicycle model-compliant paths, coupled with a grading sampling strategy to enhance computational real-time performance. Additionally, it optimizes the selection of turning points using the artificial potential field (APF) method. Compared with existing approaches, the proposed method exhibits superior performance in terms of computational efficiency and trajectory smoothness. To verify its effectiveness, both simulation experiments and puncture experiments were conducted, with a focus on analyzing computational time and the shape of actual paths. Among three experimental groups, the maximum terminal needle tip error is 1.41±0.32 mm (mean±standard deviation). The proposed algorithm can generate efficient safety paths, and it can be applied to the puncture procedures in the future.