Field-Based Spatial Self-Registration of Multicoil Hardware for B₀ Field Control

Area of Science:

• Magnetic Resonance Imaging (MRI)
• Biophysics
• Medical Imaging Technology

Background:

• Accurate B0 field control is crucial for high-resolution MRI.
• Multicoil (MC) shim hardware placement can vary between sessions, affecting B0 field homogeneity.
• Existing methods for hardware localization may require external tracking or frequent recalibration.

Purpose of the Study:

• To develop and validate a magnetic field-based self-registration method for MC shim hardware.
• To enable accurate B0 field control despite non-reproducible hardware placement.
• To improve the reproducibility of MC field control in MRI systems.

Main Methods:

• An algorithm was developed to estimate rigid transformations between reference and measured 3D field maps.
• Hardware misplacement was determined by co-registering field maps within a region of interest.
• Simulations and experimental validation with a 48-channel MC array at 3T quantified localization accuracy under varying SNR levels. A genetic algorithm optimized coil currents.

Main Results:

• Submillimeter (≤0.5mm) and subdegree (≤0.5°) localization accuracy was achieved at practical SNR levels (≥5).
• Localization errors decreased with increasing SNR and field complexity.
• The optimized field resulted in minimal translation (0.20mm) and rotation (0.11°) errors, with computation under 10 seconds.

Conclusions:

• The proposed field-based self-registration method allows rapid and accurate localization of MC shim hardware using only MR-acquired field maps.
• This eliminates the need for external tracking hardware or repeated calibrations.
• The method enhances B0 shimming performance and reproducibility in insert-based or repositionable MR systems.