Abstract
Non-invasive evaluation of functional connectivity, based on source-reconstructedestimates of phase-difference-based metrics, is notoriously non-robust. This isdue to a combination of factors, ranging from a misspecification of seed regionsto suboptimal baseline assumptions, and residual signal leakage. In this work,we propose a new analysis scheme of source-level phase-difference-basedconnectivity, which is aimed at optimizing the detection of interacting brainregions. Our approach is based on the combined use of sensor subsampling anddual-source beamformer estimation of all-to-all connectivity on a prespecifieddipolar grid. First, a pairwise two-dipole model, to account for reciprocalleakage in the estimation of the localized signals, allows for a usableapproximation of the pairwise bias in connectivity due to residual leakage of"third party" noise. Secondly, using sensor array subsampling, therecreation of multiple connectivity maps using different subsets of sensorsallows for the identification of consistent spatially localized peaks in the6-dimensional connectivity maps, indicative of true brain region interactions.These steps are combined with the subtraction of null coherence estimates toobtain the final coherence maps. With extensive simulations, we compareddifferent analysis schemes for their detection rate of connected dipoles, as afunction of signal-to-noise ratio, phase difference, and connection strength. Wedemonstrate superiority of the proposed analysis scheme in comparison tosingle-dipole models, or an approach that discards the zero phase differencecomponent of the connectivity. We conclude that the proposed pipeline allows fora more robust identification of functional connectivity in experimental data,opening up new possibilities to study brain networks with mechanisticallyinspired connectivity measures in cognition and in the clinic.