Spatial-Jitter Model for Magnetoencephalography Sensor Arrays.

Sampling jitter, i.e., random deviations in the time instants when samples are taken, causes frequency-dependent noise that reduces signal-to-noise ratio (SNR). This paper generalizes the concept of jitter to magnetoencephalography (MEG) sensor arrays that spatially sample the quasistatic magnetic field due to brain activity. It is shown that spatial jitter, i.e., random deviations in MEG sensor positions, causes spatial-frequency-dependent noise in the vector spherical harmonics domain that reduces the attainable SNR and spatial resolution in MEG. Similarly, the paper also considers noise due to random sensor orientation errors ('orientation jitter') and errors due to field integration by the finite-sized sensors ('aperture error'). The analysis in this paper shows that on-scalp MEG measurements taken closer to the head are more resistant to spatial and orientation jitter at high spatial frequencies than off-scalp measurements taken further away. On the other hand, on-scalp measurements are affected more by aperture errors than off-scalp measurements. The paper also provides new insights to the effect of sensor noise on the spatial resolution of on- and off-scalp sensor arrays using a novel normalization of the vector spherical harmonics. The paper also simulates spatial-jitter phenomena with realistic sensor arrays based on optically pumped magnetometers and superconducting quantum interference device sensors. This realistic simulation shows that spatial jitter reduces SNR and affects how the measurements should be regularized in order to maximize SNR.