Advancement in Sim-to-Real, Modeling, and AI-Driven Control for Medical Soft/Continuum Robotics

Medical soft/continuum robotics have emerged as transformative tools for minimally invasive surgery, as their inherent flexibility enables safe navigation within constrained, delicate anatomical cavities inaccessible to rigid surgical robots. However, widespread clinical translation faces a severe sim-to-real domain gap that uniquely plagues soft robotic systems. Virtual simulation platforms rely on oversimplified continuum kinematics and uniform soft-tissue material properties, failing to replicate real intraoperative complexities: heterogeneous patient tissue biomechanics, unpredictable tool-tissue contact forces, multimodal medical imaging noise, and dynamic structural deformation of compliant robotic bodies. While AI control algorithms show promise in simulated environments, they lack reliable cross-domain generalization, leading to unstable motion tracking, positioning inaccuracies, and safety risks that hinder the deployment of intelligent soft robotic systems in clinical operating rooms.This research topic aims to resolve the critical sim-to-real transfer barrier and advance integrated modeling plus AI-driven control pipelines tailored exclusively for medical soft/continuum robots. The core unsolved challenge is the poor generalization of simulation-trained control logic when deployed on physical soft robotic hardware, stemming from mismatched continuum deformation dynamics, unmodeled tissue-robot interactions, and intraoperative imaging artifacts. To tackle this limitation, we seek innovative unified frameworks that combine high-fidelity patient-specific soft tissue modeling, sim-to-real domain adaptation strategies, and closed-loop AI control guided by real-time medical imaging. Key targets include data-efficient deformation modeling, robust vision-feedback adaptive control, safety-constrained autonomous manipulation, and validated cross-domain transfer workflows. The ultimate objective is to deliver clinically viable intelligent soft robotic systems that improve precision, reduce surgical trauma, and expand autonomous minimally invasive intervention capabilities.This Research Topic welcomes original research articles, and reviews on the integration of artificial intelligence in soft/ continuum medical robotics and related medical devices. Topics of interest include, but are not limited to:(1) Sim-to-real domain adaptation for medical soft/continuum robots; (2) High-fidelity kinematic/biomechanical modeling of compliant medical manipulators;(3) Multimodal medical imaging-guided and closed-loop control of soft robotics; (4) Force-aware autonomous soft/continuum robotic manipulation for medical usage;(5) AI-driven visual servo control and reasoning of soft/continuum endoscopic robot;(6) Surgical planning and validation through transluminal intervention of soft robotics;Submissions must include the innovation in physical/simulated robot, phantom, cadaver or in-vivo testing to demonstrate sim-to-real transfer performance. Relevant clinical use cases include endoscopic, interventional, and minimally invasive targeted therapy procedures leveraging soft/continuum robotic devices. Topic editors: Zhengyang Li, Di Wu, Hangjie MO, Ge Fang Source: https://www.frontiersin.org/research-topics/83617/advancement-in-sim-to-real-modeling-and-ai-driven-control-for-medical-softcontinuum-robotics