MRI-guided endovascular robotics: A roadmap to MRI-native systems

Magnetic resonance imaging (MRI) offers potential advantages for endovascular intervention, including excellent soft-tissue contrast, multi-planar visualization, and the possibility of radiation-free guidance. Emerging robotic platforms, MRI-compatible catheter systems, and MRI-native navigation frameworks indicate that MRI-guided endovascular robotics is moving beyond isolated feasibility concepts. Yet routine MRI-guided robotic endovascular procedures remain rare. This gap is not explained by the absence of a single enabling technology, but by the interaction of tightly coupled constraints spanning instruments, imaging and feedback, navigation, and system integration. In endovascular settings, these constraints are amplified by the need for miniaturized, flexible, and steerable tools operating under limited and phase-dependent feedback. This perspective argues that progress in MRI-guided endovascular robotics requires moving beyond compatibility-driven adaptation toward MRI-robot co-design. First, the emerging landscape of representative robotic platforms, MRI-compatible and MRI-actuated catheter systems, and MRI-navigation-enabling technologies is outlined. The coupled constraints that continue to limit translation are then analyzed, showing why component-level solutions often remain fragile when integrated into full procedural workflows. Building on this framework, four research axes toward MRI-native endovascular robotic systems are proposed: MRI-native instruments, MRI-native feedback, navigation under bandwidth-limited imaging, and integrated systems built through workflow-aware validation and benchmarking. Across these axes, the discussion emphasizes the importance of miniaturized actuation, robot-aware MRI pipelines, feedback-efficient shared control, realistic simulation, benchmarking, and workflow-aware validation. MRI-guided endovascular robotics should therefore be understood as a systems integration challenge rather than a single-device problem. MRI-native robotic systems offer a concrete pathway toward safer, more precise, and radiation-free endovascular intervention.