A compact low-power magnetic particle imaging scanner based on a permanent-magnet field-free-line generator with high gradient

Field-free-line (FFL) configurations have the potential to enhance the sensitivity of magnetic particle imaging (MPI). However, electromagnet-based implementations typically suffer from high power consumption, limiting scalability toward human-sized systems, while permanent-magnet-based approaches often require large volume and weight to generate high-gradient and uniform FFLs. In this work, we present a compact permanent-magnet assembly that generates a uniform, high-gradient FFL without power consumption. The system employs four magnet pairs to produce a selection field gradient of up to 4.068 T/m/μ0 while maintaining a reduced footprint. To reduce system complexity and enrich spectral content, a single-coil, coaxial multi-frequency excitation scheme is adopted. In addition, a multi-angle FFL rotation imaging strategy combined with a system-matrix-based reconstruction framework is implemented to enhance spatial encoding and reconstruction fidelity. Magnetic field calibration confirms the generation of a stable and uniform FFL using magnet pairs with dimensions of 48 × 100 × 150 mm3. Comparative simulations demonstrate that, relative to conventional permanent-magnet designs, the proposed configuration reduces magnet volume by one order of magnitude, while completely eliminating power consumption compared to electromagnet-based systems requiring kilowatt-level power. Imaging experiments further demonstrate a spatial resolution of 0.3 mm along the main gradient direction and an imaging sensitivity of 98 ng, with reconstruction quality improving as angular coverage increases. Finally, two fully acquired multi-angle FFL system matrices and corresponding phantom datasets are open-sourced. Overall, this work provides a practical hardware solution for high-performance FFL-MPI and offers benchmark datasets to support algorithm development and system optimization toward clinically relevant MPI applications.