Peak-projection algorithm to target the phase of cortical oscillations in real-time

Abstract

Focal transcranial stimulation targeting different phases of a locally expressed oscillation can reveal the functional significance of the phase itself and may reduce the variability of the brain response to stimulation. To target the phase online, it is necessary to deal with distortions at the edge of the time-window used for online filtering and to predict the targeted phase at the time of stimulation. We propose a simple algorithm involving a padding step to minimize the filter-induced distortion and the prediction of the phase from the last peaks of the filtered signal. The novel peak-projection algorithm was validated for online phase targeting of precentral alpha activity on a set of resting-state EEG data in four healthy volunteers.