Structure-Aware and Energy-Minimized Actuation Design for Magnetic Continuum Robot Control in Vascular Intervention

Magnetic continuum robots offer a promising alternative to conventional guidewires for endovascular interventions, owing to their potential for remote and radiation-free manipulation. However, enabling closed-loop control of magnetic continuum robots within complex vascular environments remains challenging due to the anatomical intricacy, frequent directional transitions, and the need for context-aware magnetic actuation under physiological conditions. In this paper, a structure-aware and energy-minimized actuation framework (SEMAF) that consists of a structure-aware path planning module, an energy-minimized modeling module, and a magnetic actuation module is proposed. The SEMAF establishes a mapping from anatomical perception, such as bifurcation geometry, to actuation-level control parameters, including joint angles and magnetic moments. Extensive experiments, including ex vivo interventions on the right common carotid artery, cerebral aneurysms, and the middle cerebral artery, as well as in vivo studies on rabbits, validate the SEMAF’s adaptability, accuracy, and clinical potential. Experimental results demonstrate that SEMAF reduces operation time by 87.98% compared to manual guidewires and by 56.23% relative to the centerline-based intervention, while achieving an average orientation error of 10.17° and maintaining successful navigation across three representative vascular scenarios and an in vivo rabbit superior mesenteric artery intervention. These results demonstrate the feasibility of closed-loop magnetic robot navigation for next-generation minimally invasive surgery. Note to Practitioners—Conventional guidewire-based endovascular interventions expose clinicians to radiation and require extensive manual skill. Magnetic robots provide a radiation-free and remotely operable alternative, but their use in complex vascular anatomies has been limited by the lack of effective closed-loop control. This work proposes a structure-aware and energy-minimized actuation framework that maps anatomical perception, such as bifurcation geometry, to actuation-level control parameters for magnetic continuum robots. Validated through ex vivo vascular models and in vivo rabbit studies, the proposed framework achieved shorter operation times, higher navigation accuracy, and greater adaptability than manual and existing control methods. For practitioners, this framework demonstrates a feasible pathway toward safer, more efficient, and autonomous magnetic navigation in next-generation minimally invasive vascular surgery.