Research On Electrical Activities Of Cerebral Neuron With The Finite Element Method

It is regarded as an effective method to determine electrical activity sources by analysis of the brain scalp potential distribution using Electroencephalogram (EEG). Under normal circumstance, the measured scalp EEG potential shows certain regularity, however, when there is brain tissue lesions or functional changes, the corresponding regularity will be destroyed, for instance, the alteration of potential waveform. Simulating the electric activity originated with the excited nerve cell by means of finite element simulation model, working out electric potential on the surface of head model, and studying the optimized algorithm for using EEG data to get the information of equivalent dipole sources, thus it is helpful to assist the diagnosis and clinical treatment of epilepsy and other brain diseases, while also contributes to more deeply study the mechanism of the brain on the basis of electric potential on the scalp surface, this is the research objectives of this text.There are two key problems in the field of EEG, EEG forward problem and EEG inverse problem, the study of EEG forward problem is an important basis for solving the inverse problem, and whether EEG forward problem is calculated correctly or not, directly affects the site-specific accuracy. This paper conducts a simulation study on EEG forward problem with finite element method using symmetrical sphere and three-concentric-sphere model to analyze the potential distribution of the brain scalp surface. Based on the electromagnetism theory, electric potentials of the globe’s surface are obtained under the action of a certain dipole, which proves the simulation models’effectiveness, but contrast the simulation results of three-concentric-sphere model to symmetrical sphere, the former shows a good approximation. In addition, the effects of dipole parameters on the potential distribution of scalp surface are further explored in the article, such as position, direction and strength, and these differences under the same path are analyzed, as an experience for future EEG forward problem solution of each head model.Finally, the article deduces the electrical activity of neuron according to the measured scalp potential, the reasonable objective function is determined based on the best-fit between calculated and observed values of the scalp potential. In order to get more stable and the unique solution of the minimum error, the optimization of objective function under regularization processing is attained with a combination of particle swarm optimization. By comparison the observed values and the inversed ones before and after the change of the objective function, the results show that the localization precision of the source has greatly been improved than before.