Theory and Properties of Lead Field Synthesis Analysis

Abstract

Background brain activity that is unrelated to that of the brain centers under study is often referred to as brain noise. This brain noise has posed a barrier to the measurement of brain activity arising from specific sites of interest within the brain. Signal averaging, which attenuates brain noise, is limited in use to measurement of brain activity evoked by repetitive stimuli. With present MEG and EEG technology it is brain noise, not sensor noise, that limits the ability of individual MEG and EEG sensors to discriminate spontaneous neural activity. For many studies, the ideal MEG or EEG sensor required should respond to one specific region of the brain, only, as do invasive depth electrodes. To address this need, we have developed a signal analysis process for combining simultaneous measurements made from multiple physical sensors into a synthetic virtual sensor having spatially selective and directional properties. This technique, termed lead field synthesis (LFS), was verified by computer model and with actual evoked field MEG measurements.