A unified view on beamformers for M/EEG source reconstruction

• DOI: 10.1016/j.neuroimage.2021.118789
• Publication Date: 2021-12-07T07:30:49Z
• Authors: Britta U. Westner; Sarang S. Dalal; Alexandre Gramfort; Vladimir Litvak; John C. Mosher; Robert Oostenveld; Jan-Mathijs Schof...
• Keywords: Magnetoencephalography; Adaptive beamformer; Rank (graph theory)

Beamforming is a popular method for functional source reconstruction using magnetoencephalography (MEG) and electroencephalography (EEG) data. Beamformers, which were first proposed MEG more than two decades ago, have since been applied in hundreds of studies, demonstrating that they are versatile robust tool neuroscience. However, certain characteristics beamformers remain somewhat elusive there currently does not exist unified documentation the mathematical underpinnings computational subtleties as implemented most widely used academic open software packages analysis (Brainstorm, FieldTrip, MNE, SPM). Here, we provide such aims at providing background beamforming unifying terminology. Beamformer implementations compared across toolboxes pitfalls analyses discussed. Specifically, details on handling rank deficient covariance matrices, prewhitening, reduction forward fields, combination heterogeneous sensor types, magnetometers gradiometers. The overall aim this paper to contribute contemporary efforts towards higher levels transparency neuroimaging.