A Surgical Robotic System for Coronary Interventional Procedures

Robotic technologies are increasingly being explored for coronary interventions to improve procedural standardization, reduce operator burden, and enhance the consistency of device manipulation under fluoroscopic guidance. However, practical deployment remains challenging because robotic systems must continuously estimate the spatial configuration of interventional devices, respond to changing catheterization conditions, and interpret angiographic information during the procedure. In this study, we develop a robotic platform for coronary interventional procedures that integrates device localization, procedural state estimation, adaptive manipulation, and image-based analysis within a unified workflow. A probabilistic spatial inference framework is introduced to recover the three-dimensional configuration of interventional instruments from intraoperative observations, enabling accurate geometric tracking during manipulation. To support stable device advancement, a closed-loop control framework is further established to identify interaction conditions during delivery and adjust robotic actions in response to dynamic procedural changes. In parallel, learning-based angiographic image analysis is incorporated to support lesion-related assessment and intraoperative visual interpretation. Experimental validation in vascular phantoms and living animal models demonstrates reliable robotic execution across representative procedural subtasks, with favorable precision, robustness, and operational consistency. These results support the feasibility of robotic assistance for more standardized and reproducible coronary interventional procedures.