Evolutionary Automatic Guidance Scheme for Magnetic Nanoparticles in High-Flow Vascular Models Using a Uniform Magnetic Force Field
Boyoung Son, Metin Sitti, Jungwon Yoon
AI summary
Problem
Precise navigation of magnetic nanoparticles in high-flow vascular environments is hindered by dominant drag forces, loss of swarm cohesion, and the impracticality of real-time feedback. Conventional magnetic steering methods lack spatial uniformity and struggle with flow-induced particle dispersion.
Approach
The authors designed an optimized Halbach array to generate a nearly uniform magnetic force field that aligns nanoparticle chains with the flow direction. A physics-based simulator and evolutionary algorithm automatically compute robust feedforward control sequences to handle variations in chain length and injection conditions.
Key results
- Optimized hybrid dipole-quadrupole Halbach array for spatially uniform magnetic force generation
- Developed a Vascular Magnetic Particle Chain simulator modeling magnetic, drag, and wall forces
- Implemented CMA-ES evolutionary optimization to generate robust feedforward steering commands
- Experimentally validated reliable chain guidance in a four-channel vascular model, more than doubling target branch success rates
Why it matters
Establishes a scalable, feedback-free navigation strategy for precise nanoparticle delivery in dynamic vascular environments, advancing targeted drug delivery and minimally invasive therapies.
Abstract
Magnetic nanoparticle (MNP) guidance has attracted considerable attention for biomedical applications, such as targeted drug delivery and minimally invasive therapy. However, precise navigation in vivo remains challenging, particularly in high-flow vascular environments, where drag forces dominate particle dynamics and real-time feedback is impractical. Here, we present an evolutionary automatic guidance scheme for feedforward control of MNP chains in a vascular model. The proposed approach leverages chain alignment with the flow direction to achieve directional migration into specific branches, without relying on swarm cohesion or online feedback. To provide uniform actuation, a Halbach array is designed and optimized to generate a nearly uniform magnetic force field within the target workspace. A physics-based simulator incorporating magnetic, drag, and wall interaction forces is developed to model chain dynamics, and control sequences are optimized using the Covariance Matrix Adaptation Evolution Strategy (CMA-ES), ensuring robustness to variations in chain length and injection conditions. The method is experimentally validated using a four-channel vascular model, demonstrating that feedforward magnetic actuation can reliably guide nanoparticle chains under physiologically relevant high-flow conditions. This study establishes a practical and scalable strategy for nanoparticle navigation, providing a foundation for future biomedical applications in dynamic vascular environments.