Adaptive Localization Method for SERF Atomic Magnetometers Under Magnetic Field Disturbances

Spin-exchange relaxation-free (SERF) atomic magnetometers (AMs) offer ultrahigh sensitivity and noncryogenic operation, making them ideal for wearable neurophysiological monitoring. However, their requirement for a near-zero magnetic field makes them susceptible to static magnetic disturbances induced by human movement. These disturbances degrade sensitivity and trigger nonlinear responses, generating higher-order harmonics that severely compromise neural source localization accuracy. Against this backdrop, this article establishes an atomic spin dynamics model to theoretically reveal how static magnetic disturbances generate higher-order harmonics. Subsequently, numerical simulations analyze the impact of multidirectional static fields on sensor response and measurement accuracy. A harmonic-feature-based adaptive localization method is then presented. By evaluating the harmonic distortion in individual sensor outputs, HALM adaptively adjusts their weights during neural source localization, effectively mitigating static magnetic interference. The algorithm's effectiveness is validated using a computational current dipole model in a simulated interference environment. Ultimately, this proof-of-concept study provides a useful technical reference for the future application of SERF-based magnetoencephalography (MEG) systems in complex magnetic field environments.