Environment-Driven Synthetic Baseline Analysis and Optimization in Joint Measurement OPM-MEG Arrays.

Optically pumped magnetometer-based magnetoencephalography (OPM-MEG), with its flexible sensor configuration and wide range of application scenarios, has become a powerful complement to conventional superconducting quantum interference device magnetoencephalography (SQUID-MEG). However, this higher flexibility also means that OPM-MEG sensor arrays are more susceptible to interference from complex and variable background magnetic noise. Previous research has shown that deploying reference sensors around the scalp array for noise cancellation is an effective strategy. Nonetheless, the selection of its key parameter, the spatial distance between the reference and scalp sensors, commonly termed the synthetic baseline, predominantly relies on empirical rules and lacks systematic theoretical optimization. To address this issue, this study thoroughly investigates the fundamental impact of the synthetic baseline on the system's noise suppression performance. Simulation results demonstrate that the optimal baseline length is not a fixed value but varies systematically with environmental noise characteristics and the specific requirements of the source localization task. Building on this analysis, a Baseline Adaptive Reference Optimization (BARO) method is proposed. As an environment-driven strategy, the BARO method automatically determines the optimal baseline configuration by maximizing the output signal-to-noise ratio (SNR). Compared to traditional fixed-baseline configurations, the proposed BARO method significantly enhances the output SNR and effectively reduces the localization error of equivalent current dipoles within the brain across various simulated complex noise scenarios. This work provides a physically interpretable criterion for baseline optimization and offers theoretical support for environment-adaptive configuration of OPM-MEG sensor arrays.