RESEARCH ARTICLE Anatomically constrained minimum variance beamforming applied to EEG Vyacheslav Murzin • Armin Fuchs • J. A. Scott Kelso Received: 17 September 2010 / Accepted: 20 August 2011 / Published online: 14 September 2011 Ó Springer-Verlag 2011 Abstract Neural activity as measured non-invasively using electroencephalography (EEG) or magnetoencepha- lography (MEG) originates in the cortical gray matter. In the cortex, pyramidal cells are organized in columns and acti- vated coherently, leading to current flow perpendicular to the cortical surface. In recent years, beamforming algo- rithms have been developed, which use this property as an anatomical constraint for the locations and directions of potential sources in MEG data analysis. Here, we extend this work to EEG recordings, which require a more sophisticated forward model due to the blurring of the electric current at tissue boundaries where the conductivity changes. Using CT scans, we create a realistic three-layer head model consist- ing of tessellated surfaces that represent the cerebrospinal fluid-skull, skull-scalp, and scalp-air boundaries. The cor- tical gray matter surface, the anatomical constraint for the source dipoles, is extracted from MRI scans. EEG beam- forming is implemented on simulated sets of EEG data for three different head models: single spherical, multi-shell spherical, and multi-shell realistic. Using the same condi- tions for simulated EEG and MEG data, it is shown (and quantified by receiver operating characteristic analysis) that EEG beamforming detects radially oriented sources, to which MEG lacks sensitivity. By merging several tech- niques, such as linearly constrained minimum variance beamforming, realistic geometry forward solutions, and cortical constraints, we demonstrate it is possible to localize and estimate the dynamics of dipolar and spatially extended (distributed) sources of neural activity. Keywords Electroencephalography  EEG  Source reconstruction  Inverse problem  Beamforming  Boundary element method Introduction Non-invasive techniques such as electroencephalography (EEG), magnetoencephalography (MEG), magnetic reso- nance imaging (MRI), functional MRI (fMRI), positron emission tomography (PET), and computed tomography (CT) provide complementary measures of structure, func- tion, and dynamics of the living human brain and have proved essential for scientific research and clinical diag- nostics. The zoo of imaging technologies is based on dif- ferent physical principles and exploits different tissue properties and physiological responses. MRI and fMRI measure the level of water and the hemodynamic response (changes in blood flow and level of oxygen in the blood due to both intrinsic and task-induced properties of the brain), respectively. The neural sources in the brain pro- duce electric currents whose effects can be measured in the form of magnetic fields outside the head (MEG) or electric potentials on the scalp surface (EEG). These measurements have the common primary goal of estimating the locations of neural activity inside the brain. Functional MRI, when compared to M/EEG, offers real 3-D imaging with high spatial resolution but due to the nature of the hemodynamic V. Murzin (&)  A. Fuchs  J. A. S. Kelso Center for Complex Systems and Brain Sciences, Florida Atlantic University, Boca Raton, FL 33431, USA e-mail: murzin@ccs.fau.edu A. Fuchs Physics Department, Florida Atlantic University, Boca Raton, FL 33431, USA J. A. S. Kelso Intelligent Systems Research Centre, University of Ulster, Derry, Northern Ireland, UK 123 Exp Brain Res (2011) 214:515–528 DOI 10.1007/s00221-011-2850-5