Learning Expert Strategy for Autonomous Robotic Endovascular Intervention via Decoupled Procedural Execution

Jul 16, 2026

Upload by Jiven

Endovascular interventions are high-stakes procedures demanding precise device control inside complex, tortuous vascular anatomy. Autonomous endovascular navigation can standardize procedural quality and reduce variability, yet existing reinforcement learning approaches often lack explicit constraint satisfaction and safety guarantees. The document proposes a learning-based expert strategy that decouples high-level strategic decision-making from low-level procedural execution. A strategic RL policy generates global navigation intents, then an expert-informed execution module enforces expert norms, real-time kinematic limits, and vessel safety constraints, validated in simulation and a real robotic platform.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages8

Reading Time20 min

Abstract

Introduction

Learning Expert Strategy for Autonomous Robotic Endovascular Intervention via Decoupled Procedural Execution Yanxi Chen 1 ,2 , Tianliang Yao3 , Shaolong Tang 1 , Jiyuan Zhao 1 , Hengyu Hu 1 ,4 , Zhaoxing Li 1 , Antonio Snchez Egea5 , Peng Qi 1 ,∗ arXiv :2607 .00066v1 [ cs .RO] 30 Jun 2026 Abstract—Endovascular interventions are high-stakes procedures requiring precise device operation within complex and tortuous vascular anatomies. Autonomous endovascular navigation has the potential to standardize procedural quality and reduce the performance variability inherent in manual operation. Although Reinforcement Learning (RL) approaches have demonstrated promise in enabling autonomy in endovascular intervention, they often struggle with explicit constraint satisfaction and safety guarantees. To address these challenges, a learning-based expert strategy is introduced, enhancing procedural consistency in autonomous endovascular intervention by explicitly decoupling high-level strategic decision-making from low-level procedural execution. The proposed framework replicates the expert clinical decision-making process: a strategic RL policy generates global navigation intents, which are subsequently refined through an expert-informed execution module. This module ensures that robot movements strictly adhere to expert operational norms, real-time kinematic limits, and vessel safety constraints. Experimental evaluation across high-fidelity 3D simulations and a real-world robotic platform demonstrates that the proposed framework not only outperforms baseline policies but also effectively replicates expertlevel proficiency. The framework achieves a high navigation success rate (> 96%) and a 29.3% reduction in operational steps, which translates to enhanced operative efficiency and minimized device-vessel interaction. Furthermore, a 13% reduction in trajectory variance indicates superior procedural standardization, aligning autonomous behavior with established clinical norms. These results underscore its potential to enhance the predictability, safety, and consistency of robotic endovascular interventions. I. INTRODUCTION Endovascular interventions are essential minimally invasive procedures for treating cardiovascular and neurovas This work has been accepted by IEEE/RSJ IROS 2026 . Copyright maybe transferred without notice, after which this version may no longer be accessible. This work is supported by the National Key Research and Development Program of China under Grant No. 2023YFB4705200, and the National Natural Science Foundation of China under Grant No. 52575034. The authors would like to thank Mr. Tao Liu from Shanghai Operation Robot Co., Ltd., for providing technical support in experiments. (*Corresponding Author: Peng Qi, [email: pqi@tongji.edu.cn](email: pqi@tongji.edu.cn)). 1Department of Control Science and Engineering, College of Electronicsand Information Engineering, and Shanghai Institute of Intelligent Science and Technology, Tongji University, Shanghai 200092, China; 2 School of Mechanical Engineering, Tongji University, Shanghai 200092, China; 3Department of Electronic Engineering, Faculty of Engineering, The Chinese University of Hong Kong, Hong Kong SAR 999077, China; 4 School of Mechatronic Engineering and Automation, Shanghai University, Shanghai 200444, China; 5Department of Mechanical Engineering, Universitat Politcnica de Catalunya (UPC), Barcelona 08034, Spain. Fig. 1. The robot-assisted endovascular interventions are the minimally invasive procedures that utilize interventional guidewires, a C-arm Xray machine, and endovascular robotics. (a) Manual operation: clinicians directly operate the robotic system based on their clinical expertise and procedural experience. (b) Autonomous operation: an automated console generates automatic instructions based on real-time fluoroscopes to endovascular robotics, where clinicians supervise the procedure. The lower panel illustrates guidewire navig

FAQs

What key problem does the proposed expert strategy address in autonomous endovascular intervention?

It targets the difficulty reinforcement learning methods have in explicitly satisfying constraints and providing safety guarantees during endovascular procedures.

How does the framework decouple decision-making from execution?

A strategic RL policy produces global navigation intents, which are then refined by an expert-informed execution module that applies expert operational norms and safety constraints.

What performance improvements are reported in the evaluation?

Experiments show a navigation success rate above 96%, a 29.3% reduction in operational steps, and a 13% reduction in trajectory variance, indicating more standardized and efficient procedures.

Jul 16, 2026

Upload by Jiven