P 59. Real-time prefrontal beta phase estimation during emotional interference inhibition

Abstract

Previous studies suggest a common brain network including the right inferior frontal gyrus and pre-supplementary motor area integral to various types of emotional interference inhibition, i.e. suppression of task-irrelevant stimuli to pursue goal-directed behavior in the context of emotional distraction. This network reacts differently to emotional interference in individual subjects and exhibits a specific spectral signature in the beta-band. Here we analyze 64 channel surface EEG data during emotional interference inhibition operationalized by a Flanker task with preceding emotionally salient stimuli to determine “true phase” estimates and perform real-time (predictive) phase estimation of task-related beta-band oscillations. EEG data was spectrally analyzed, signal-to-noise ratios (SNRs) of periodic components of the power spectra calculated, and non-causal phase estimations (i.e. using data extending both directions around the time point of interest) performed to yield “true phase” estimates of task-related beta oscillations after Flanker signal presentation. Causal phase estimation of task-related EEG beta oscillations was achieved using an autoregressive forward prediction algorithm on data preceding the time point of interest only. Results show that real-time (i.e. causal) phase estimation of beta oscillations in the emotional Flanker task is feasible with satisfactory circular deviations of “true phase” estimates, but needs individual optimization of phase estimation parameters (i.e. window size [samples], FIR filter order, number of samples removed at data window edge, order of the autoregressive model). Findings have implications for the design of real-time EEG-triggered brain oscillation-synchronized TMS experiments to causally study the role of task-related beta-band activity for emotional interference inhibition. This study was supported by the Boehringer Ingelheim Foundation.“