Novel vector beamformers for EEG source imaging

Abstract

This paper introduces two novel vector beamforming algorithms, namely the Vector Weight Normalized and Vector Standardized Minimum Variance beamformers, for brain source localization and reconstruction. Our mathematical analysis shows that the Vector Weight Normalized Minimum Variance beamformer (V-WNMVB) is the true vector version of the Synthetic Aperture Magnetoencephalography (SAM). Our Monte-Carlo simulation results with fixed and rotated dipole sources show that the two new vector beamformers give better source localization errors than the existing ones, including SAM, linearly constrained minimum variance beamformer and vector Borgiotti-Kaplan beamformer. Finally, the multiple dipole source simulation studies show that the performance of V-WNMVB is as good as that of SAM, however, it does not require any assumption on the source orientation. © 2011 IEEE.