Low Resolution Brain Source Localization Using EEG Signals

Abstract

Each mental or physical task gives rise to generate electromagnetic activity in the brain. These electrical signals are analyzed by using various neuroimaging techniques which include electroencephalography (EEG), magnetoencephalogy (MEG), positron emission tomography (PET) and functional magnetic resonance imaging (fMRI). However, when the brain sources which are responsible for such electrical activity are localized, then it’s called brain source localization or source estimation. This information is utilized to comprehend brain’s physiological, pathological, mental, functional abnormalities. Also, the information is used to diagnose cognitive behaviour of the brain. Various methodologies based upon EEG signals are adopted to localize the active sources such as minimum norm estimation (MNE), low resolution brain electromagnetic tomography (LORETA), standardized LORETA, exact LORETA, multiple signal classification (MUSIC), focal underdetermined system solution (FOCUSS) etc. This research discusses localizing ability of low resolution techniques (LORETA and sLORETA) for various head models (finite difference model and concentric model). The simulations are carried out by using NETSTATION software. The results are compared in terms of activations for same EEG data with the same stimulus provided to subjects. However, it is observed that the combination of finite difference method (FDM) with sLORETA produced best results in terms of source intensity level (nA). Hence, the combination of inverse method sLORETA with FDM produces better source localization.