Research Analyzer
← Back ICRA 2026

Fluoroscopy-Constrained Magnetic Robot Control Via Zernike-Based Field Modeling and Nonlinear MPC

Xinhao Chen, Hongkun Yao, Anuruddha Bhattacharjee, Suraj Raval, Lamar Mair, Yancy Diaz-Mercado, Axel Krieger

PDF

AI summary

Key figure (auto-extracted from paper)
A unified NMPC controller combined with Zernike field modeling and a Kalman filter enables precise magnetic robot navigation despite low-frame-rate, noisy fluoroscopic feedback.
Magnetic robot control Nonlinear MPC Zernike polynomials Fluoroscopy Drug delivery Kalman filter

Problem

Clinical magnetic robot control is hindered by fluoroscopy's low frame rate and noisy pose tracking, which degrade conventional high-frequency control methods.

Approach

The authors develop a direct NMPC framework that outputs coil currents using an analytically differentiable Zernike polynomial magnetic field model and a Kalman filter to estimate robot states from sparse, noisy feedback.

Key results

  • Zernike polynomial model provides high-fidelity, analytically differentiable magnetic field approximation
  • NMPC with Kalman filter maintains high accuracy at 3 Hz feedback with 2 mm Gaussian noise
  • Achieves 1.18 mm RMS position error in spine phantom drug delivery while respecting safety boundaries
  • Outperforms baseline two-layer and lookup-table control methods under fluoroscopy-mimicking conditions

Why it matters

Makes magnetic surgical robots clinically viable by ensuring precise, safe navigation under real-world fluoroscopic imaging constraints.

Abstract

Magnetic actuation enables surgical robots to navigate complex anatomical pathways while reducing tissue trauma and improving surgical precision. However, clinical deployment is limited by the challenges of controlling such systems under fluoroscopic imaging, which provides low frame rate and noisy pose feedback. This paper presents a con- trol framework that remains accurate and stable under such conditions by combining a nonlinear model predictive con- trol (NMPC) framework that directly outputs coil currents, an analytically differentiable magnetic field model based on Zernike polynomials, and a Kalman filter to estimate the robot state. Experimental validation is conducted with two magnetic robots in a 3D-printed fluid workspace and a spine phantom replicating drug delivery in the epidural space. Results show the proposed control method remains highly accurate when feedback is downsampled to 3 Hz with added Gaussian noise (σ = 2 mm), mimicking clinical fluoroscopy. In the spine phantom experiments, the proposed method successfully executed a drug delivery trajectory with a root mean square (RMS) position error of 1.18 mm while maintaining safe clearance from critical anatomical boundaries.

Index terms

Medical Robots and Systems Motion Control Micro/Nano Robots

Related papers