Eeg, Meg Is The Numerical Solution Of The Different Brain Model Comparison

EEG(Electroencephalography) and MEG(Magnetoenphalography) are harmless and noninva-sive medicine imaging technologies for investigating neuronal activity in the living human brain. They have been widely used in diverse areas including the normal working brain, brain diseases diagnosis, and the dynamics of neural networks studies, etc. The computing of EEG/MEG for-ward problem is required in the research of localization of current dipole source (DSL) in the brain. The EEG forward problem computes the scalp potentials from the current sources in the brain and the MEG forward problem solves the external magnetic field from the current sources.In previous work, numerical simulations were performed in one head model(centric sphere) for testing their methods. To our knowledge, the comparison between the effects of nonspherical model and sphere model on the forward solution of EEG/MEG is limited. In this paper, we use two head models:ovoid head model and sphere head model. The comparison of the maps for scalp potentials, the field of outer magnetic and their relative errors are made between two head models for different position and direction of the dipole source parameters. We investigate the influence of different head models and dipole source parameters on the scalp potentials and external magnetic. The results show that the patterns of potentials map in different head models are similar. However, the magnitude of potential and the field of outer magnetic is significantly influenced by the head model. Dipole location has great effects on the peak value of potential. The previous methods used in computing the forward problem are boundary element method, finite difference method and finite element method. The disadvantages are the high degree of freedom, low computing efficiency and the inflexibility of computer programs. We develop algorithms based on BEM of Weighted Residuals which has advantages of low degree of freedom, high computing efficiency and the flexibility of computer programs. The method of BEM often used constant or linear basis functions. In this work, we use constant and orthogonal basis functions.This paper is organized as follows. In chapter 1, we make a simple introduction of the physi-ological basis, measuring instrument and the definition of forward problem about EEG and MEG. In chapter 2, we induce the field equation of EEG/MEG forward problem and give the definition of head model which we used to solve the forward problem. In chapter 3, BEM based on Weighted Residuals is introduced and the algorithms of forward problem is constructed. In chapter 4, we perform the numerical simulations in two head models, analyze the simulation results and obtain the influence of the different head models and dipole source parameters on the forward EEG/MEG problems.